The Chennai Mathematical Institute (CMI) MSc Data Science programme is a rigorous, 64-credit, four-semester degree designed to transition quantitative graduates into advanced predictive analytics and machine learning roles. The curriculum equips students with a strong foundation in mathematics, statistics, and computing, progressing to specialized coursework in distributed systems, algorithmic design, and advanced statistical modeling, alongside a compulsory summer internship. Eligibility is deliberately broad: candidates must hold an undergraduate degree (B.A., B.Sc., B.Math., B.Stat., B.E.,
Algebra & Number Theory
Functions & Calculus
Sets, Logic & Relations
Unit 1 — Discrete Mathematics
Combinatorics & Induction
Probability & Random Variables
Statistics & Data Analysis
Algorithmic Thinking
Algebra & Number Theory
Functions & Calculus
Sets, Logic & Relations
Unit 1 — Discrete Mathematics
Combinatorics & Induction
Probability & Random Variables
Statistics & Data Analysis
Algorithmic Thinking
Algebra & Number Theory
Functions & Calculus
Sets, Logic & Relations
Unit 1 — Discrete Mathematics
Combinatorics & Induction
Probability & Random Variables
Statistics & Data Analysis
Algorithmic Thinking
Algebra & Number Theory
Functions & Calculus
Sets, Logic & Relations
Unit 1 — Discrete Mathematics
Combinatorics & Induction
Probability & Random Variables
Statistics & Data Analysis
Algorithmic Thinking
Q1. (CAT 2023) Find the limit \[ \lim_{x\to\infty}\left(x-x\cos\frac{1}{\sqrt{x}}\right) \]
Answer: none
Solution: Insight: This is an $\infty - \infty$ indeterminate form at infinity; factoring $x$ and substituting $t = 1/\sqrt{x}$ converts it to the standard limit $(1-\cos t)/t^2 = 1/2$.
Exam route:
1. Factor out $x$: $x(1 - \cos(1/\sqrt{x}))$.
2. Substitute $t = 1/\sqrt{x}$, so $t \to 0^+$ and $x = 1/t^2$.
3. The expression becomes $(1 - \cos t)/t^2$.
4. Apply the standard limit $\lim_{t\to 0} (1-\cos t)/t^2 = 1/2$.
Learning route:
This is a limits-at-infinity question with an $\infty - \infty$ indeterminate form, recognisable because both $x$ and $x\cos(1/\sqrt{x})$ grow without bound as $x \to \infty$. The trigger for the substitution method is the $1/\sqrt{x}$ inside the cosine, which shrinks to zero as $x$ grows.
Step 1: Never split $\infty - \infty$ into two separate limits. Instead, factor out $x$:
$$x - x\cos\frac{1}{\sqrt{x}} = x\left(1 - \cos\frac{1}{\sqrt{x}}\right)$$
This converts the form from $\infty - \infty$ to $\infty \cdot 0$, which is still indeterminate but easier to handle.
Step 2: Substitute $t = 1/\sqrt{x}$. As $x \to \infty$, $t \to 0^+$, and $x = 1/t^2$:
$$x\left(1 - \cos\frac{1}{\sqrt{x}}\right) = \frac{1}{t^2}(1 - \cos t) = \frac{1 - \cos t}{t^2}$$
Step 3: Apply the standard limit. Using $1 - \cos t = 2\sin^2(t/2)$ and $\lim_{u\to 0}\frac{\sin u}{u} = 1$:
$$\lim_{t\to 0}\frac{1-\cos t}{t^2} = \lim_{t\to 0}\frac{2\sin^2(t/2)}{t^2} = \lim_{t\to 0}\frac{2\cdot(t/2)^2}{t^2} = \frac{1}{2}$$
Wrong path: A tempting mistake is to split the limit: $\lim_{x\to\infty} x - \lim_{x\to\infty} x\cos(1/\sqrt{x})$. This yields $\infty - \infty$, which is invalid because you can only split limits if both individual limits exist and are finite.
Q2. (CAT 2020) Let \(f(x)\) be a real-valued function all of whose derivatives exist. Recall that a point \(x_0\) in the domain is called an <b>inflection point</b> of \(f(x)\) if the second derivative \(f''(x)\) changes sign at \(x_0\). Given the function \[ f(x)=\frac{x^5}{20}-\frac{x^4}{2}+3x+1, \] which of the following statements are true?
Answer: ["A","B"]
Solution: Key idea: This is a "statement_truth" question requiring the systematic identification and verification of inflection points by checking the sign change of the second derivative.
Step 1: Find the first derivative. $f'(x) = \frac{d}{dx}\left(\frac{x^5}{20} - \frac{x^4}{2} + 3x + 1\right) = \frac{x^4}{4} - 2x^3 + 3$.
Step 2: Find the second derivative. $f''(x) = \frac{d}{dx}\left(\frac{x^4}{4} - 2x^3 + 3\right) = x^3 - 6x^2$.
Step 3: Find candidate points where $f''(x) = 0$. $x^3 - 6x^2 = x^2(x - 6) = 0 \implies x = 0$ or $x = 6$.
Step 4: Verify sign change at $x = 0$. For $x < 0$, $x^2 > 0$ and $x - 6 < 0$, so $f''(x) < 0$. For $0 < x < 6$, $x^2 > 0$ and $x - 6 < 0$, so $f''(x) < 0$. Since the sign does not change, $x = 0$ is NOT an inflection point.
Step 5: Verify sign change at $x = 6$. For $0 < x < 6$, $f''(x) < 0$. For $x > 6$, $x^2 > 0$ and $x - 6 > 0$, so $f''(x) > 0$. Since the sign changes from negative to positive, $x = 6$ IS an inflection point.
Step 6: Evaluate options. Option A is true ($x_0=0$ is not an inflection point). Option B is true ($x_0=6$ is the only one). Options C and D are false.
Answer: Options A and B.
Q3. (CAT 2021) For a non-zero real number \(a\), the inverse of \(J=\begin{pmatrix}a&1&0\\0&a&1\\0&0&a\end{pmatrix}\) is
Answer: ["B"]
Solution: Key idea: This is a matrix inverse problem for a special upper triangular matrix, recognizable because it has a constant diagonal $a$ and constant superdiagonal $1$.
Step 1: Decompose the matrix as $J = aI + N$, where $N = \begin{pmatrix} 0 & 1 & 0 \\ 0 & 0 & 1 \\ 0 & 0 & 0 \end{pmatrix}$.
Step 2: Observe that $N$ is nilpotent. Specifically, $N^2 = \begin{pmatrix} 0 & 0 & 1 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \end{pmatrix}$ and $N^3 = 0$.
Step 3: Use the finite geometric series expansion for the inverse: $(aI + N)^{-1} = a^{-1}(I + a^{-1}N)^{-1} = a^{-1}(I - a^{-1}N + a^{-2}N^2)$.
Step 4: Substitute $I$, $N$, and $N^2$ into the expansion:
$J^{-1} = \begin{pmatrix} a^{-1} & 0 & 0 \\ 0 & a^{-1} & 0 \\ 0 & 0 & a^{-1} \end{pmatrix} - \begin{pmatrix} 0 & a^{-2} & 0 \\ 0 & 0 & a^{-2} \\ 0 & 0 & 0 \end{pmatrix} + \begin{pmatrix} 0 & 0 & a^{-3} \\ 0 & 0 & 0 \\ 0 & 0 & 0 \end{pmatrix} = \begin{pmatrix} a^{-1} & -a^{-2} & a^{-3} \\ 0 & a^{-1} & -a^{-2} \\ 0 & 0 & a^{-1} \end{pmatrix}$.
Answer: B
Q4. (CAT 2024) Starting with the number \(n=1\), we generate a sequence of numbers. In the second step we replace 1 by either \(2=2\times 1\) or \(3=2\times 1+1\). In general, replace \(n\) by either \(2n\) or \(2n+1\) to get a new value for \(n\). Which of the following numbers can be obtained as values of \(n\) in this fashion?
Answer: ["A","C"]
Solution: Key idea: The operations $n \to 2n$ and $n \to 2n+1$ correspond to appending a binary digit ($0$ or $1$) to the right of the binary representation of $n$.
Since we start with $n=1$ (which is $(1)_2$), any number generated by this process will have a binary representation that starts with $1$ and consists only of the digits appended during the steps. Crucially, since every positive integer has a unique binary representation starting with $1$, **every positive integer** can be generated by this process.
Let's verify this for each option by converting to binary and checking if it can be reached from $1$ by appending bits.
Option A: 7
$7 = (111)_2$.
Start: $1 = (1)_2$.
Step 1: Append 1 $\to 2(1)+1 = 3 = (11)_2$.
Step 2: Append 1 $\to 2(3)+1 = 7 = (111)_2$.
So, 7 is obtainable.
Option B: 10
$10 = (1010)_2$.
Start: $1 = (1)_2$.
Step 1: Append 0 $\to 2(1) = 2 = (10)_2$.
Step 2: Append 1 $\to 2(2)+1 = 5 = (101)_2$.
Step 3: Append 0 $\to 2(5) = 10 = (1010)_2$.
So, 10 is obtainable.
Option C: 15
$15 = (1111)_2$.
Start: $1 = (1)_2$.
Step 1: Append 1 $\to 3 = (11)_2$.
Step 2: Append 1 $\to 7 = (111)_2$.
Step 3: Append 1 $\to 15 = (1111)_2$.
So, 15 is obtainable.
Option D: 2026
Any positive integer $N$ can be written in binary. The process of generating $N$ from $1$ corresponds exactly to reading the binary digits of $N$ from left to right (excluding the leading 1 which is our start state) and applying $2n$ for '0' and $2n+1$ for '1'.
Since 2026 is a positive integer, it has a binary representation.
$2026 = 1024 + 512 + 256 + 128 + 64 + 32 + 8 + 2 = (11111101010)_2$.
It starts with 1. We can reach it by following the bits after the first one.
So, 2026 is obtainable.
Wait, let me re-read the question carefully. "Which of the following numbers can be obtained...?"
Usually, in such MSQ questions, if all are correct, all should be selected. Let's double check if there's a constraint I missed.
"Starting with n=1... replace n by either 2n or 2n+1".
This generates the set of all positive integers.
Proof: By strong induction. Base case: 1 is in the set. Assume all integers $< k$ are in the set. If $k$ is even, $k=2m$, then $m < k$, so $m$ is in the set, and we can get $k$ from $m$ by $2m$. If $k$ is odd, $k=2m+1$, then $m < k$, so $m$ is in the set, and we can get $k$ from $m$ by $2m+1$. Thus all positive integers are reachable.
Therefore, 7, 10, 15, and 2026 are all obtainable.
Answer: A, B, C, D
Q5. (CAT 2024) Which of the following statements is/are true for real numbers \(x,y\)?
Answer: ["B","D"]
Solution: Key idea: This is a universal statement evaluation question involving real number properties. We must test each implication for all real numbers $x, y$, looking for counterexamples to disprove false statements.
Step 1: Analyze Option A: If $x^2 = y^2$ then $x = y$.
Counterexample: Let $x = 1$ and $y = -1$. Then $1^2 = (-1)^2 = 1$, but $1 \neq -1$. Thus, $x = \pm y$. Statement A is False.
Step 2: Analyze Option B: If $x^3 = y^3$ then $x = y$.
The function $f(t) = t^3$ is strictly increasing for all real $t$. Therefore, it is one-to-one (injective). If $x^3 = y^3$, taking the cube root of both sides yields $x = y$. Statement B is True.
Step 3: Analyze Option C: If $x < y$ then $x^2 < y^2$.
Counterexample: Let $x = -2$ and $y = 1$. Then $-2 < 1$, but $(-2)^2 = 4$ and $1^2 = 1$. Here $4 > 1$, so $x^2 > y^2$. Statement C is False.
Step 4: Analyze Option D: If $x < y$ then $x^3 < y^3$.
Since $f(t) = t^3$ is strictly increasing on $\mathbb{R}$, $x < y$ implies $f(x) < f(y)$, i.e., $x^3 < y^3$. Statement D is True.
Answer: Options B and D are true.
Q1. (CAT 2020) How many squares are there on a \(7\times 7\) chessboard?
Answer: ["D"]
Solution: Key idea: this is a grid enumeration question asking for the total number of squares of all sizes in an $n \times n$ grid. The trigger is "how many squares", which implies counting $1 \times 1$, $2 \times 2$, up to $n \times n$ squares, not just the unit cells.
Step 1: A $k \times k$ square on a $7 \times 7$ board is uniquely determined by the position of its top-left corner.
Step 2: The top-left corner can be placed in $(7 - k + 1) = (8 - k)$ horizontal positions and $(8 - k)$ vertical positions. Thus, there are $(8 - k)^2$ squares of size $k \times k$.
Step 3: Sum over all possible sizes $k$ from 1 to 7:
Total squares $= \sum_{k=1}^{7} (8-k)^2 = 7^2 + 6^2 + 5^2 + 4^2 + 3^2 + 2^2 + 1^2$.
Step 4: Calculate the sum: $49 + 36 + 25 + 16 + 9 + 4 + 1 = 140$. (This matches the standard formula $\frac{n(n+1)(2n+1)}{6}$ for $n=7$, which gives $\frac{7 \times 8 \times 15}{6} = 140$).
Step 5: Match with options. Option A (49) counts only the $1 \times 1$ cells. Option B (204) is the sum for an $8 \times 8$ board. Option D (140) is the correct total for a $7 \times 7$ board.
Answer: ["D"]
Q2. (CAT 2022) A relation \(R\) on the set \(A=\{a,b,c,d\}\) is defined by reading the columns of the following table from top to bottom. If a column in the table reads \((x,y,1)\) it means \(x\) is related to \(y\) in \(R\). If a column in the table reads \((x,y,0)\) it means \(x\) is not related to \(y\).<br/><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 560 95" width="560" height="95"><rect x="30" y="10" width="500" height="70" fill="white" stroke="black"/><line x1="30" y1="35" x2="530" y2="35" stroke="black"/><line x1="30" y1="58" x2="530" y2="58" stroke="black"/><line x1="155" y1="10" x2="155" y2="80" stroke="black"/><line x1="280" y1="10" x2="280" y2="80" stroke="black"/><line x1="405" y1="10" x2="405" y2="80" stroke="black"/><g font-size="15" text-anchor="middle" font-family="serif"><text x="45" y="28">a</text><text x="75" y="28">b</text><text x="105" y="28">c</text><text x="135" y="28">d</text><text x="170" y="28">a</text><text x="200" y="28">b</text><text x="230" y="28">c</text><text x="260" y="28">d</text><text x="295" y="28">a</text><text x="325" y="28">b</text><text x="355" y="28">c</text><text x="385" y="28">d</text><text x="420" y="28">a</text><text x="450" y="28">b</text><text x="480" y="28">c</text><text x="510" y="28">d</text><text x="45" y="51">a</text><text x="75" y="51">a</text><text x="105" y="51">a</text><text x="135" y="51">a</text><text x="170" y="51">b</text><text x="200" y="51">b</text><text x="230" y="51">b</text><text x="260" y="51">b</text><text x="295" y="51">c</text><text x="325" y="51">c</text><text x="355" y="51">c</text><text x="385" y="51">c</text><text x="420" y="51">d</text><text x="450" y="51">d</text><text x="480" y="51">d</text><text x="510" y="51">d</text><text x="45" y="74">0</text><text x="75" y="74">0</text><text x="105" y="74">1</text><text x="135" y="74">0</text><text x="170" y="74">1</text><text x="200" y="74">0</text><text x="230" y="74">0</text><text x="260" y="74">0</text><text x="295" y="74">1</text><text x="325" y="74">1</text><text x="355" y="74">0</text><text x="385" y="74">1</text><text x="420" y="74">0</text><text x="450" y="74">0</text><text x="480" y="74">1</text><text x="510" y="74">0</text></g></svg><br/>For instance, from the fifth column we have, \((a,b)\in R\), and from the second we have \((b,a)\notin R\).<br/>Another relation \(S\) on the set \(A\) is defined as: for any \(x,y\in A\), the pair \((x,y)\) is in \(S\) if and only if there exists \(z\in A\) such that both \((x,z)\in R\) and \((z,y)\in R\) hold.<br/>Which of the following pairs are in \(S\)?
Answer: ["A","C"]
Solution: Key idea: This is a **Relation Composition via Matrix/Table** problem. Recognise it because relation $S$ is defined as "exists $z$ such that $(x,z) \in R$ and $(z,y) \in R$", which is precisely the definition of $R^2$ or $R \circ R$.
Step 1: Decode the table into relation $R$.
The table columns represent pairs $(row\_label, col\_label, value)$. Reading the bottom row (values) against the top two rows (labels):
Col 1: $(a,a,0)$
Col 2: $(b,a,0)$
Col 3: $(c,a,1) \implies (c,a) \in R$
Col 4: $(d,a,0)$
Col 5: $(a,b,1) \implies (a,b) \in R$
Col 6: $(b,b,0)$
Col 7: $(c,b,0)$
Col 8: $(d,b,0)$
Col 9: $(a,c,1) \implies (a,c) \in R$
Col 10: $(b,c,1) \implies (b,c) \in R$
Col 11: $(c,c,0)$
Col 12: $(d,c,1) \implies (d,c) \in R$
Col 13: $(a,d,0)$
Col 14: $(b,d,0)$
Col 15: $(c,d,0)$
Col 16: $(d,d,0)$
So $R = \{(c,a), (a,b), (a,c), (b,c), (d,c)\}$.
Step 2: Compute $S = R \circ R$.
We need pairs $(x,y)$ where a path $x \to z \to y$ exists in $R$.
Let's trace paths of length 2 from each starting node:
- From $a$:
$a \to b \to c$ (since $(a,b) \in R, (b,c) \in R$) $\implies (a,c) \in S$
$a \to c \to a$ (since $(a,c) \in R, (c,a) \in R$) $\implies (a,a) \in S$
- From $b$:
$b \to c \to a$ (since $(b,c) \in R, (c,a) \in R$) $\implies (b,a) \in S$
- From $c$:
$c \to a \to b$ (since $(c,a) \in R, (a,b) \in R$) $\implies (c,b) \in S$
$c \to a \to c$ (since $(c,a) \in R, (a,c) \in R$) $\implies (c,c) \in S$
- From $d$:
$d \to c \to a$ (since $(d,c) \in R, (c,a) \in R$) $\implies (d,a) \in S$
So $S = \{(a,c), (a,a), (b,a), (c,b), (c,c), (d,a)\}$.
Step 3: Check options against $S$.
A: $(a,a) \in S$ — TRUE
B: $(b,b) \notin S$ — FALSE
C: $(c,c) \in S$ — TRUE
D: $(d,d) \notin S$ — FALSE
Answer: Options A and C are correct.
Q3. (CAT 2020) It is mid-semester exam week at CMI and first-year students from both M.Sc. Data Science (DS) and M.Sc. Computer Science (CS) have their exams scheduled for Monday from 10 a.m. to 1 p.m. in Lecture Hall 1. The first row in Lecture Hall 1 has six seats. In how many different ways can three M.Sc. DS students - Anish, Binish and Finish - and three M.Sc. CS students - Ramesh, Suresh, and Ragesh - be seated in this row, in such a way that two students from the same course do not sit next to each other?
Answer: ["C"]
Solution: Key idea: This is an alternating arrangement problem, recognizable by the condition "two students from the same course do not sit next to each other".
Step 1: We have 3 DS students and 3 CS students. To ensure no two students from the same course sit together, they must strictly alternate.
Step 2: There are exactly two valid alternating patterns for 6 seats:
Pattern 1: DS - CS - DS - CS - DS - CS
Pattern 2: CS - DS - CS - DS - CS - DS
Step 3: For Pattern 1, the 3 DS students can be arranged in their 3 seats in $3! = 6$ ways. The 3 CS students can be arranged in their 3 seats in $3! = 6$ ways. Total for Pattern 1 = $6 \times 6 = 36$.
Step 4: Similarly, for Pattern 2, the arrangements = $3! \times 3! = 36$.
Step 5: Total valid arrangements = $36 + 36 = 72$.
Answer: 72
Q4. (CAT 2023) A perfect shuffle of a deck of cards divides the deck into two equal parts and then interleaves the cards from each half, starting with the first card of the first half.<br/>For instance, if we shuffle a deck of cards containing 10 cards arranged \([1,2,3,4,5,6,7,8,9,10]\), we first create two equal decks with cards \([1,2,3,4,5]\) and \([6,7,8,9,10]\) and then interleave them to get a new deck \([1,6,2,7,3,8,4,9,5,10]\).<br/>We start with the deck \([8,1,4,5,3,6,2,7]\) and keep shuffling. Which card(s) will never appear next to 5?
Answer: ["D"]
Solution: Key idea: This is a permutation cycle decomposition question, recognizable because it asks about the long-term behavior of a deterministic rearrangement (shuffle).
Step 1: Understand the shuffle mapping. For an 8-card deck, the perfect shuffle maps positions as follows: $p \to 2p-1$ if $p \le 4$, and $p \to 2(p-4)$ if $p > 4$.
Step 2: Find the cycles of positions.
- Position 1 maps to 1. (Cycle: {1})
- Position 8 maps to 8. (Cycle: {8})
- Position 2 $\to$ 3 $\to$ 5 $\to$ 2. (Cycle: {2, 3, 5})
- Position 4 $\to$ 7 $\to$ 6 $\to$ 4. (Cycle: {4, 6, 7})
Step 3: Track the cards in these cycles.
Initial deck: `[8, 1, 4, 5, 3, 6, 2, 7]` at positions 1 to 8.
- Cycle {1} always contains card 8.
- Cycle {8} always contains card 7.
- Cycle {2, 3, 5} contains cards {1, 4, 3}.
- Cycle {4, 6, 7} contains cards {5, 6, 2}.
Step 4: Determine adjacencies for card 5. Card 5 is always in Cycle {4, 6, 7}.
- If 5 is at pos 4, neighbors are pos 3 and 5 (both in Cycle {2, 3, 5}, cards {1, 3, 4}).
- If 5 is at pos 6, neighbors are pos 5 (Cycle {2, 3, 5}) and pos 7 (Cycle {4, 6, 7}).
- If 5 is at pos 7, neighbors are pos 6 (Cycle {4, 6, 7}) and pos 8 (Cycle {8}, card 7).
Step 5: Check the options. Card 8 is fixed at position 1. For 8 to be next to 5, 5 would need to be at position 2. However, 5 only visits positions 4, 6, and 7. Thus, 8 can never be next to 5.
Answer: 8
Q5. (CAT 2022) A ternary tree starts with a single root node at the top of the tree. Each node in the tree can have up to three nodes as its children. No node in the tree is the child of two different nodes. A node which has no children is called a leaf node.<br/>The children of a node are drawn below it, connected by edges. The level of a node \(v\) in the ternary tree is the number of edges in the (unique) path from the root node to \(v\). Thus, for instance, the root node is at level 0, and each child of the root node is at level 1.<br/>A complete ternary tree is a ternary tree in which (i) each non-leaf node has exactly three children, and (ii) all leaf nodes are at the same level. This latter level is called the height of the complete ternary tree. The complete ternary trees of heights 0, 1, and 2, respectively are shown in the figure below, where we use the symbol \(\otimes\) to denote a node.<br/><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 560 180" width="560" height="180"><g fill="none" stroke="black" stroke-width="1.5"><text x="70" y="65" font-size="20" text-anchor="middle">⊗</text><text x="205" y="35" font-size="20" text-anchor="middle">⊗</text><line x1="205" y1="45" x2="165" y2="95"/><line x1="205" y1="45" x2="205" y2="95"/><line x1="205" y1="45" x2="245" y2="95"/><text x="165" y="115" font-size="20" text-anchor="middle">⊗</text><text x="205" y="115" font-size="20" text-anchor="middle">⊗</text><text x="245" y="115" font-size="20" text-anchor="middle">⊗</text><text x="405" y="25" font-size="20" text-anchor="middle">⊗</text><line x1="405" y1="35" x2="325" y2="85"/><line x1="405" y1="35" x2="405" y2="85"/><line x1="405" y1="35" x2="485" y2="85"/><text x="325" y="105" font-size="20" text-anchor="middle">⊗</text><text x="405" y="105" font-size="20" text-anchor="middle">⊗</text><text x="485" y="105" font-size="20" text-anchor="middle">⊗</text><line x1="325" y1="110" x2="295" y2="145"/><line x1="325" y1="110" x2="325" y2="145"/><line x1="325" y1="110" x2="355" y2="145"/><line x1="405" y1="110" x2="375" y2="145"/><line x1="405" y1="110" x2="405" y2="145"/><line x1="405" y1="110" x2="435" y2="145"/><line x1="485" y1="110" x2="455" y2="145"/><line x1="485" y1="110" x2="485" y2="145"/><line x1="485" y1="110" x2="515" y2="145"/></g><g font-size="20" text-anchor="middle"><text x="295" y="165">⊗</text><text x="325" y="165">⊗</text><text x="355" y="165">⊗</text><text x="375" y="165">⊗</text><text x="405" y="165">⊗</text><text x="435" y="165">⊗</text><text x="455" y="165">⊗</text><text x="485" y="165">⊗</text><text x="515" y="165">⊗</text></g></svg><br/>What is the total number of nodes in a complete ternary tree of height 9?
Answer: ["C"]
Solution: Key idea: The total number of nodes in a complete $k$-ary tree of height $h$ is the sum of a geometric progression.
Step 1: Identify the number of nodes at each level. In a complete ternary tree, level $k$ has exactly $3^k$ nodes.
Step 2: The tree has height 9, meaning the levels range from $k=0$ (the root) to $k=9$.
Step 3: The total number of nodes is the sum of nodes at all levels: $\sum_{k=0}^{9} 3^k$.
Step 4: Apply the geometric series sum formula $S_n = \frac{a(r^n - 1)}{r - 1}$, where $a=1$, $r=3$, and the number of terms is $10$ (from 0 to 9).
Step 5: Calculate the sum: $\frac{1 \cdot (3^{10} - 1)}{3 - 1} = \frac{3^{10}-1}{2}$.
Answer: Option C.
Q1. (CAT 2021) Fifteen telephones are received at a service center. Of these, 5 are mobile, 6 are cordless, and 4 are wired. These 15 phones are randomly numbered from 1 to 15 to establish the order in which they are serviced. Which of the following statement(s) is/are correct?
Answer: ["A","B"]
Solution: Key idea: this is a *sampling without replacement* question with a *random ordering* of 15 items split into three types (5M, 6C, 4W). The probability of any ordered type-pattern is computed by multiplying the changing fractions, or equivalently by counting favourable permutations against total permutations.
**Option A.** "First and third are mobile, second is not."
- P(1st is M) = 5/15.
- Given that, 14 phones remain, 10 of which are not M. P(2nd is not M) = 10/14.
- Given that, 13 phones remain, 4 of which are M. P(3rd is M) = 4/13.
- Product = (5·10·4)/(15·14·13). Option A is **correct**.
**Option B.** "First four serviced are all wired."
- P = (4/15)·(3/14)·(2/13)·(1/12) = 4!/(15·14·13·12) = 24/32760 = 1/1365.
- Also 1/ C(15,4) = 1/1365 (choosing which 4 positions the 4 wired phones occupy among the first 4 is forced). Option B is **correct**.
**Option C.** "After servicing 10 phones, only one type remains."
- The 5 unserviced phones must all be of one type. Only M has 5 phones, so the 5 unserviced must be exactly the 5 mobiles.
- P = C(5,5)/C(15,5) = 1/3003. Option C gives C(6,5)/C(15,5) = 6/3003, which is wrong (6 cordless cannot fit into 5 slots). Option C is **wrong**.
**Option D.** "Two of each type among the first six."
- Favourable count = C(5,2)·C(6,2)·C(4,2) (choose 2 M, 2 C, 2 W independently), divided by C(15,6).
- Option D writes a sum in the numerator instead of a product. Option D is **wrong**.
Answer: A, B.
Q2. (CAT 2021) Common Description: <b>Description for following two questions:</b> In the 2019-2020 season of the English Premier League (EPL), 380 matches were played in a home and away format. The figure below describes the number of goals scored by the home team and the away team against the number of matches played. For example, the home team scored one goal in 125 matches.<br/><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 760 300" width="760" height="300"><g fill="none" stroke="black" stroke-width="1"><line x1="60" y1="240" x2="310" y2="240"/><line x1="60" y1="240" x2="60" y2="35"/><line x1="450" y1="240" x2="700" y2="240"/><line x1="450" y1="240" x2="450" y2="35"/></g><g fill="white" stroke="black"><rect x="85" y="139" width="28" height="101"/><rect x="125" y="91" width="28" height="149"/><rect x="165" y="122" width="28" height="118"/><rect x="205" y="184" width="28" height="56"/><rect x="245" y="221" width="28" height="19"/><rect x="285" y="230" width="20" height="10"/><rect x="475" y="93" width="28" height="147"/><rect x="515" y="87" width="28" height="153"/><rect x="555" y="142" width="28" height="98"/><rect x="595" y="201" width="28" height="39"/><rect x="635" y="230" width="20" height="10"/><rect x="665" y="235" width="16" height="5"/><rect x="690" y="238" width="12" height="2"/></g><g font-size="11" font-family="serif" text-anchor="middle"><text x="99" y="135">84</text><text x="139" y="87">125</text><text x="179" y="118">99</text><text x="219" y="180">47</text><text x="259" y="217">16</text><text x="295" y="226">8</text><text x="489" y="89">123</text><text x="529" y="83">128</text><text x="569" y="138">82</text><text x="609" y="197">33</text><text x="645" y="226">8</text><text x="673" y="231">4</text><text x="696" y="234">1</text><text x="185" y="285">(a)</text><text x="575" y="285">(b)</text><text x="185" y="270">Number of goals scored by the home team</text><text x="575" y="270">Number of goals scored by the away team</text></g></svg> Which of the following statement(s) is/are correct?
Answer: ["A","B","C"]
Solution: Insight: This is an empirical probability question, recognizable because it provides frequency data from a bar chart and asks you to verify statements about proportions and cumulative percentages.
Exam route: Extract frequencies from the chart for the specific conditions in each option, compute the cumulative sums, and divide by the total matches $N = 380$ to check the claimed percentages.
Learning route:
Step 1: Extract the data. The problem states $N = 380$ matches.
Home team goals: 0:84, 1:125, 2:99, 3:47, 4:16, 5:8. (Sum = 379, we use $N=380$ as the denominator per the problem statement).
Away team goals: 0:123, 1:128, 2:82, 3:33, 4:8, 5:4, 6:1. (Sum = 379).
Step 2: Verify Statement A. "Home team scored at most one goal" means 0 or 1 goal.
Count = $84 + 125 = 209$.
Percentage = $\frac{209}{380} \times 100 \approx 55.0\%$. Since $55.0\% > 50\%$, Statement A is TRUE.
Step 3: Verify Statement B. "Away team scored more than two goals" means 3, 4, 5, or 6 goals.
Count = $33 + 8 + 4 + 1 = 46$.
Percentage = $\frac{46}{380} \times 100 \approx 12.1\%$. Since $12.1\% > 10\%$, Statement B is TRUE.
Step 4: Verify Statement C. "Home team scored three or more goals" means 3, 4, or 5 goals.
Count = $47 + 16 + 8 = 71$.
Percentage = $\frac{71}{380} \times 100 \approx 18.7\%$. Since $18.7\% > 15\%$, Statement C is TRUE.
Step 5: Verify Statement D. "Away team scored less than three goals" means 0, 1, or 2 goals.
Count = $123 + 128 + 82 = 333$.
Percentage = $\frac{333}{380} \times 100 \approx 87.6\%$. Since $87.6\%$ is NOT more than $90\%$, Statement D is FALSE.
Final Answer: A, B, and C are correct.
Q3. (CAT 2021) Common Description: Description for following two questions: A Non-Banking Finance Corporation (NBFC) declares fixed annual rates of simple interest on their auto and housing loans each year. The rates of interest offered by the company differ from year to year depending on the variation in macro economic indicators like inflation, RBI’s repo rate etc. The annual rates of interest offered by the company for the Auto and Housing sectors over the years are shown in the figure. <br/> <svg xmlns="http://www.w3.org/2000/svg" width="520" height="300" viewBox="0 0 520 300"> <rect x="0" y="0" width="520" height="300" fill="white"/> <line x1="80" y1="230" x2="430" y2="230" stroke="black"/> <line x1="80" y1="50" x2="80" y2="230" stroke="black"/> <text x="70" y="232" text-anchor="end" font-size="10">0</text> <text x="70" y="172" text-anchor="end" font-size="10">5</text> <text x="70" y="112" text-anchor="end" font-size="10">10</text> <text x="70" y="52" text-anchor="end" font-size="10">15</text> <rect x="115" y="98" width="13" height="132" fill="white" stroke="black"/> <line x1="116" y1="227" x2="127" y2="216" stroke="black"/> <line x1="116" y1="214" x2="127" y2="203" stroke="black"/> <line x1="116" y1="201" x2="127" y2="190" stroke="black"/> <line x1="116" y1="188" x2="127" y2="177" stroke="black"/> <line x1="116" y1="175" x2="127" y2="164" stroke="black"/> <line x1="116" y1="162" x2="127" y2="151" stroke="black"/> <line x1="116" y1="149" x2="127" y2="138" stroke="black"/> <line x1="116" y1="136" x2="127" y2="125" stroke="black"/> <line x1="116" y1="123" x2="127" y2="112" stroke="black"/> <line x1="116" y1="110" x2="127" y2="99" stroke="black"/> <rect x="130" y="110" width="13" height="120" fill="white" stroke="black"/> <text x="118" y="93" font-size="10">11</text> <text x="130" y="105" font-size="10">10</text> <text x="129" y="250" text-anchor="middle" font-size="10">2010</text> <rect x="160" y="74" width="13" height="156" fill="white" s
The Chennai Mathematical Institute (CMI) MSc Data Science programme is a rigorous, 64-credit, four-semester degree designed to transition quantitative graduates into advanced predictive analytics and machine learning roles. The curriculum equips students with a strong foundation in mathematics, statistics, and computing, progressing to specialized coursework in distributed systems, algorithmic design, and advanced statistical modeling, alongside a compulsory summer internship.
Eligibility is deliberately broad: candidates must hold an undergraduate degree (B.A., B.Sc., B.Math., B.Stat., B.E., B.Tech., or equivalent) with a background in Mathematics, Statistics, or Computer Science. This explicitly welcomes engineers, physicists, and final-year undergraduates, dispelling the myth that only pure mathematics or computer science degrees are accepted.
Admission is determined by a dedicated, independent entrance exam held annually since 2018. Running for 3.5 hours (2:00 PM to 5:30 PM) across roughly 37 national centers, this offline, pen-and-paper test features 40 questions totaling 100 marks. It is split into Part A (objective) and Part B (descriptive), with zero negative marking and partial credit awarded for reasoning. Crucially, the entrance syllabus-focusing on school-level mathematics, discrete mathematics, probability theory, and basic programming-is entirely distinct from the actual degree curriculum.
CMI's own placement cell publishes a year-by-year table of maximum, mean and median campus offers going back to 2014-15 - but this data is institute-wide across all of CMI's undergraduate and postgraduate programmes combined, since CMI does not release a breakdown specific to MSc Data Science alone.
| Year | Maximum Offer | Mean Offer | Median Offer |
|---|---|---|---|
| 2024-25 | ₹37.5 LPA | ₹17.7 LPA | ₹16.0 LPA |
| 2023-24 | ₹25.8 LPA | ₹18.2 LPA | ₹18.6 LPA |
| 2022-23 | ₹47 LPA | ₹20.8 LPA | ₹20.7 LPA |
| 2021-22 | ₹62 LPA | ₹17.6 LPA | ₹16 LPA |
| 2020-21 | ₹18.4 LPA | ₹12.99 LPA | ₹13.5 LPA |
| 2019-20 | ₹20 LPA | ₹14 LPA | ₹13.35 LPA |
| 2018-19 | ₹16.54 LPA | ₹12.88 LPA | ₹14.8 LPA |
Read across the decade, the trend is a broadly rising mean and median offer up to the ₹16-20 LPA range in recent years, though maximum offers swing sharply year to year (from ₹18.4 LPA in 2020-21 to ₹62 LPA in 2021-22) - a pattern consistent with a small total cohort, where a handful of exceptional offers can move the maximum without changing the typical outcome much, which is why the median is a steadier number to anchor expectations on than the maximum.
CMI's own framing is directly relevant to Data Science applicants specifically: the placement page states that MSc Data Science students, together with MSc Computer Science and BS (Hons.) Mathematics and Computer Science students, make up the largest share of participants in campus interviews each year, so this institute-wide trend is one Data Science graduates are heavily represented within, even though CMI does not isolate their figures alone.
CMI Data Science refers to the M.Sc. in Data Science offered by Chennai Mathematical Institute (CMI), a two-year, four-semester postgraduate degree that CMI awards directly under its status as a university recognised under Section 3 of the UGC Act, 1956.
Launched in 2018, the programme is built around three intersecting pillars - mathematics, statistics and computer science - with the explicit aim of training graduates for data-analytics roles in industry rather than for a purely academic track. A student needs a minimum of 64 credits (16 regular courses) to graduate, split across mathematical methods, probability and statistics with R, programming with Python, linear algebra, data mining and machine learning, distributed computing, and a compulsory three-month summer industry internship between the first and second year. The second year opens into six elective slots chosen from a bank of roughly fifteen courses spanning machine learning, finance, NLP, computer vision and optimisation, letting students build a specialisation inside the broader degree.
Eligibility is open to graduates from B.A., B.Sc., B.Math., B.Stat., B.E. or B.Tech. backgrounds with prior exposure to mathematics, statistics or computer science - so engineers, physicists and pure-science graduates are all eligible alongside statistics and CS majors. Admission for the 2026-27 batch (which began classes on 3 August 2026) ran entirely through CMI's own written entrance examination, with no CUET, JEE, or GATE route accepted. The programme runs from CMI's Siruseri campus on Chennai's OMR IT corridor.
The M.Sc. Data Science admission test is a separate, dedicated paper - not shared with any of CMI's other entrance exams - and it has run as its own distinct question paper every year since the programme's first intake in 2018, with official past papers and solutions published for 2018 through 2026.
Two features make it different from CMI's other tests. First, timing: while all of CMI's exams are held on the same afternoon, the MSc Data Science paper (like the BS paper) runs from 2:00 PM to 5:30 PM - half an hour longer than the 2:00-5:00 PM window given to the MSc/PhD Mathematics, MSc/PhD Computer Science and PhD Physics papers. Second, content emphasis: rather than the advanced Class XI-XII algebra, calculus, geometry and number theory that dominate the BS paper, or the algorithms, automata theory and mathematical logic of the MSc/PhD Computer Science paper, the Data Science paper leans on school-level mathematics, discrete mathematics, probability theory and the ability to read and trace pseudocode - reflecting the applied, data-oriented focus of the degree rather than pure mathematical depth.
A candidate who selects the MSc Data Science examination can only apply to the MSc Data Science programme in that cycle - unlike the Mathematics and Computer Science papers, which can be combined with each other or with the corresponding PhD track. Selection for MSc Data Science is also unusual among CMI's postgraduate options in that it typically does not involve an interview; interviews are convened only at the discretion of the Admissions Committee based on prior academic record, whereas MSc Mathematics candidates are always interviewed. For the 2026-27 cycle, CMI's published results list shows 65 candidates offered admission to MSc Data Science, alongside 31 to MSc Computer Science.
CMI's MSc Data Science entrance exam runs for 3.5 hours (2:00 PM to 5:30 PM) in a single national afternoon sitting, held offline in pen-and-paper mode at roughly 37 exam centres across India - half an hour longer than the 3-hour window (2:00-5:00 PM) given to CMI's MSc/PhD Mathematics, MSc/PhD Computer Science and PhD Physics papers, reflecting this paper's larger descriptive component.
The paper has 40 questions worth 100 marks total, split into two parts. Part A has 20 objective-type questions worth 2 marks each (40 marks total), with no partial credit - many are multiple-select "which of the following are true" items where every correct option must be chosen to earn the marks. Part B has 20 short-answer questions worth 3 marks each (60 marks total), where partial credit is explicitly available for correct method and justification even if the final answer is not fully correct. Based on CMI's own published sample papers and official solution keys, there is no negative marking anywhere in this scheme.
Selecting the MSc Data Science entrance paper restricts an applicant to that programme alone for the admission cycle - it cannot be combined with the Mathematics or Computer Science papers the way those two can be combined with each other. Results are released roughly a month after the exam as a selection list only; CMI does not disclose individual marks, percentile, or an all-India rank to candidates, and most selected candidates are admitted directly, with an interview called only at the Admissions Committee's discretion based on academic record.
Graduates leave with a specific, named toolkit rather than vague "analytical skills" - the curriculum is built to produce fluency in Python and R programming, SQL and relational database design, and data visualisation grounded in Tufte's data-ink and graphical-integrity framework.
On the mathematical side, students work through numerical linear algebra (LU and QR factorisation, Jacobi and Gauss-Seidel methods, singular value decomposition and PCA), classical statistical inference (maximum likelihood estimation, hypothesis testing, sampling distributions), and convex and combinatorial optimisation (linear programming, gradient and conjugate-gradient methods). The machine-learning sequence covers supervised methods (linear and logistic regression, LDA/QDA, decision trees, support vector machines), unsupervised methods (clustering, association-rule mining), and a dedicated Advanced Machine Learning course on deep neural networks, PyTorch and Keras, reinforcement learning and hidden Markov models. A distributed-computing and big-data course adds working exposure to Hadoop, Spark and MapReduce-style processing.
Depending on electives chosen, students can additionally graduate with named competencies in Bayesian data analysis (Stan, MCMC, Hamiltonian Monte Carlo), time-series forecasting (ARIMA/GARCH, Kalman filters), natural language processing, computer vision, topological data analysis, or quantitative finance (portfolio theory, financial time-series, algorithmic trading) - plus a completed three-month industry internship as applied, real-world experience.
Admission follows a fixed sequence: online application, a single written entrance exam, a merit-based selection list, and - for most Data Science candidates - direct admission without an interview.
Applications open online (at CMI's yearly apply[year].cmi.ac.in portal) in early March and close in early April; for the 2026-27 cycle the window ran from 2 March to 4 April 2026. Applicants register with an email and phone number, fill in personal and academic details, select the MSc Data Science examination specifically (this choice locks the applicant into that programme alone for the cycle), upload a photograph, signature and mark-sheets, choose a preferred test city from roughly 37 centres nationwide, and pay the application fee online. Admit cards are released about a week before the exam.
The written exam is held on a single national afternoon (2 May 2026 for the 2026-27 cycle), lasting from 2:00 PM to 5:30 PM. Results follow roughly a month later; CMI does not release marks or an all-India rank to candidates - the results page simply lists selected Applicant IDs, sorted by ID rather than merit position. For the 2026-27 cycle, 65 candidates were listed as selected for MSc Data Science. Most selected candidates are admitted directly on this basis; CMI's Admissions Committee retains discretion to call individual candidates for an interview based on academic record, but this is not a standard second stage for this programme (unlike MSc Mathematics, which always interviews). SC, ST, OBC-NCL, EWS and PwD candidates must submit the relevant certificate in CMI's prescribed format at the time of admission to claim reserved-category consideration. Confirmed candidates receive an offer letter by email, and the academic session begins in early August.
Because the MSc Data Science paper draws on school-level mathematics, discrete mathematics, probability and basic programming logic rather than the advanced pure-mathematics syllabus used for CMI's BS and MSc/PhD Mathematics papers, an effective strategy looks different from generic "CMI exam" advice - it should be built around data-interpretation and applied reasoning, not topology or real analysis.
Step 1 - Build the four foundational pillars. Work systematically through school-level algebra, matrices, determinants, logarithms, functions and elementary calculus; discrete mathematics (sets, combinatorics, the pigeonhole principle, the binomial theorem, mathematical induction, boolean logic); probability theory (conditional probability, Bayes' theorem, standard distributions, expectation and variance, summary statistics); and the ability to trace simple pseudocode with variables, loops and conditionals. CMI's own syllabus note recommends standard references such as Sheldon Ross's "A First Course in Probability" and C.L. Liu's "Elements of Discrete Mathematics" for exactly this stage.
Step 2 - Topic-wise practice. Once each pillar's basics are comfortable, drill topic-wise problem sets in each of the four areas separately, focusing especially on multi-part data-interpretation questions (bar-graph and percentage-based problems have appeared repeatedly) and "select all correct options" reasoning, since CMI's objective section rewards only fully-correct selections with no partial credit.
Step 3 - Past papers. CMI has published a full official question paper and solution set for MSc Data Science every year from 2018 through 2026 - an unusually deep and reliable practice archive for a niche exam. Work through these in chronological order, since later years show a shift toward more layered, multi-statement "which of the following are true" questions.
Step 4 - Timed mock tests. Simulate the exact format: 40 questions in 3.5 hours, split into a 20-question objective Part A (2 marks each) and a 20-question descriptive Part B (3 marks each, partial credit available). For Part B specifically, practise writing full justifications rather than only final answers, since partial credit is awarded for correct reasoning even when the final number is wrong.
Step 5 - Structured revision. Keep an error log sorted by the four topic pillars rather than by paper, and revisit recurring high-frequency topics (Bayes' theorem, matrix transformations, function properties, counting problems, and code-tracing) in the final weeks rather than starting new topics.
Suggested time allocation by starting point (Indicative, not official):
Strong background (engineering, statistics, or working data professionals): roughly 60% of prep time on Steps 3-4 (past papers and mocks), 40% on light refreshers of weaker pillars.
Moderate or mixed background (science or humanities graduates with some but not continuous exposure to maths): a more even split - 40% foundational rebuilding across all four pillars, 30% topic practice, 30% past papers and mocks.
Early-stage or rusty-fundamentals starters: front-load 55-60% of total prep time on Step 1 before touching past papers at all, since CMI's descriptive section penalises shaky fundamentals more than superficial gaps in speed.
Two different things get called the "CMI Data Science syllabus," and it is worth separating them clearly: the entrance-exam syllabus (what to study to get admitted) and the programme syllabus (what is taught once enrolled) are distinct documents covering different material.
The entrance-exam syllabus, published officially by CMI as a standalone PDF, covers four areas: school-level mathematics (progressions, means, polynomials, matrices, determinants, linear equations, number theory, logarithms, function properties, elementary calculus), discrete mathematics (sets and relations, combinatorics, the pigeonhole principle, the binomial theorem, mathematical induction, boolean logic), probability theory (conditional probability, Bayes' theorem, standard distributions, expectation, variance, data interpretation and summary statistics), and basic programming (reading and interpreting pseudocode with variables, conditionals and loops). In terms of exam weightage, this translates to a 40:60 split between Part A (objective) and Part B (descriptive) out of 100 total marks - CMI does not publish a further topic-by-topic weightage breakdown beyond this two-part split.
The programme syllabus, by contrast, is the actual two-year, four-semester course of study - detailed subject-by-subject in the subjects, units and chapters tables elsewhere on this page - running from foundational mathematics, statistics-with-R and Python programming in Semester I, through linear algebra, machine learning and big-data infrastructure in Semester II, into applied statistical learning and the first electives in Semester III, and a fully elective final semester drawn from roughly fifteen named courses. There is no separate downloadable syllabus PDF covering this full programme curriculum the way the entrance exam has one; the fullest official source for the programme syllabus is CMI's own Data Science programme pages and its 2026-27 Information Brochure, both of which this page draws on directly. Students preparing for the entrance exam should focus on the four exam-syllabus areas above rather than the broader programme curriculum, which is only relevant once admitted.
Last updated: October 08, 2026
You need targeted, topic-wise notes for the CMI MSc Data Science entrance exam because generic mathematics or computer science materials will waste your preparation time. The dedicated Data Science paper has run independently since 2018, testing a specific blend of school-level mathematics, discrete math, probability, and programming over a unique 3.5-hour window.
Your revision notes must prioritize the four core pillars that constitute over 93 percent of the 293 analyzed past year questions. Focusing your condensed notes on these areas yields the highest return on investment.
| Subject Area | Questions (out of 293) | Weightage |
|---|---|---|
| School Level Mathematics | 109 | 37.20% |
| Probability Theory | 67 | 22.87% |
| Discrete Mathematics | 62 | 21.16% |
| Programming | 35 | 11.95% |
What most candidates get wrong here: They spend equal time drafting notes for all subjects. You should allocate your note-making time proportionally. Prioritize matrices, conditional probability, combinatorics, and basic algorithmic thinking before touching low-weightage topics like advanced number theory.
Your practical next action today is to open your current notes and tag every page with one of these four categories. If a page does not fit, archive it.
Authentic preparation requires solving the distinct question papers published annually by the Chennai Mathematical Institute. Archives of these papers and their official solutions are available from 2018 through 2026 on the CMI admissions portal.
Be cautious with third-party free PDF compilations. Many are outdated, hidden behind paywalls, or incorrectly blend the Data Science syllabus with the general Mathematics or Computer Science papers, which have entirely different question styles and difficulty curves.
Access Curated PYQs with Solutions
Effective study materials depend on the exam's unique two-part structure. The paper features 40 questions worth 100 marks, with zero negative marking and partial credit awarded for descriptive reasoning in Part B.
Focus your notes on speed, pattern recognition, and elimination shortcuts.
Your notes must document step-by-step logical proofs and algorithmic tracing, not just final numerical answers, to maximize partial credit.
View Detailed Preparation Strategy
The CMI MSc Data Science entrance exam is a dedicated 3.5-hour offline paper running from 2:00 PM to 5:30 PM. It features 40 questions worth 100 marks, with Part A containing 20 objective questions. These short notes target the official 4-area entrance syllabus, backed by exact weightage from 293 previous year questions to maximize your revision efficiency. Last updated: October 09, 2026.
The exam is a distinct 100-mark, 40-question paper held offline at roughly 37 centres across India. Eligibility explicitly welcomes final-year undergraduates and diverse quantitative backgrounds including B.A., B.Sc., B.E., and B.Tech. graduates with Mathematics, Statistics, or Computer Science foundations.
| Event | Expected Schedule (IST) |
|---|---|
| Application Window | |
| Admit Card Release | |
| Entrance Examination |
Your revision must strictly follow the official entrance-exam syllabus, which covers School Level Mathematics, Discrete Mathematics, Probability Theory, and Programming. This is entirely distinct from the post-enrollment programme syllabus, and studying advanced machine learning or deep learning topics now will waste your preparation time.
Aspirants frequently confuse the entrance syllabus with the MSc curriculum. They waste weeks on advanced statistics or Python libraries instead of mastering pseudocode tracing, pigeonhole principle applications, and school-level calculus, which actually appear on the test.
Practical next action: Download the official standalone PDF syllabus from the CMI admissions page today and cross-check your current notes against these four specific areas.
Analysis of 293 previous year questions reveals that School Level Mathematics, Probability Theory, and Discrete Mathematics collectively account for over 81 percent of the exam. Prioritizing these high-yield chapters ensures maximum marks per hour of study.
| Subject Area | Questions (out of 293) | Weightage | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| School Level Mathematics | 109 | 37.20% | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Probability Theory | 67 | 22.87% |
| Category | Approx. Cutoff (Indicative) | Safe Range |
|---|---|---|
| General | Not published by CMI | [NEEDS VERIFICATION] |
| OBC-NCL | Not published by CMI | [NEEDS VERIFICATION] |
| EWS | Not published by CMI | [NEEDS VERIFICATION] |
| SC | Not published by CMI (relaxed qualifying score applies) | [NEEDS VERIFICATION] |
| ST | Not published by CMI (relaxed qualifying score applies) | [NEEDS VERIFICATION] |
| PC (PwD) | Not published by CMI (relaxed qualifying score applies) | [NEEDS VERIFICATION] |
There is effectively no rank-vs-marks relationship to analyse for CMI's MSc Data Science exam, because CMI does not compute or publish an all-India rank at all - its official results page states outright that admitted candidates are listed by application ID, not ranked, which is a fundamentally different model from JEE- or NEET-style exams where rank and marks are both disclosed and closely tracked.
In practical terms, this means a candidate cannot benchmark their preparation against a published "rank corresponding to X marks" table the way they could for a large national exam - CMI's internal selection is a closed process based on that year's applicant pool and question paper difficulty, not a fixed marks-to-rank curve. The only actionable guidance that follows from this is to maximise raw score against the paper itself: since Part A's 2-mark objective questions carry no partial credit, every fully correct answer there counts in full, while Part B's 3-mark descriptive questions reward complete, well-justified reasoning even when the final numeric answer is off.
The core academic requirement is an undergraduate degree - B.A., B.Sc., B.Math., B.Stat., B.E., B.Tech. or an equivalent - with a background in Mathematics, Statistics or Computer Science; this is CMI's own stated wording, and it is deliberately broader than a "Statistics or CS degree only" requirement, since it also admits engineers, physicists and other quantitative graduates. Final-year undergraduates who expect to complete their degree by the start of the relevant academic year are eligible to apply and appear for the entrance exam.
CMI's official brochure and admissions pages do not state a minimum qualifying percentage for MSc Data Science eligibility. [NEEDS VERIFICATION: some third-party aggregator sites cite an unverified minimum percentage (commonly around 70%) that could not be confirmed against any CMI-published source.] Similarly, no age limit and no cap on the number of exam attempts appear anywhere in CMI's official eligibility material for this programme.
Unlike CMI's BS programmes, which allow direct admission for top performers in national Mathematics and Informatics Olympiads, and its PhD programmes, which accept GATE, JEST, NBHM or UGC-CSIR NET scores as alternative qualification routes, MSc Data Science has no alternative qualification channel at all - every applicant, regardless of academic record, must sit CMI's written entrance exam.
Reservation follows Government of India policy: CMI provides proportional representation and a relaxed qualifying score for Scheduled Caste (SC), Scheduled Tribe (ST), Other Backward Classes-Non-Creamy Layer (OBC-NCL), Persons with Disabilities of 40% or more (PC), and Economically Weaker Section (EWS, family income under ₹8 lakh a year and not otherwise covered by SC/ST/OBC) candidates, who must submit the relevant certificate in CMI's prescribed format at the time of admission. CMI does not offer a management or NRI quota.
CMI Data Science graduates recruit primarily into data analytics, machine learning and quantitative roles at technology, finance and analytics firms, based on the recruiter list CMI's own placement cell publishes: Credit Suisse, Ernst & Young, Tata Research Development and Design Centre (TRDDC), Adobe, Zendrive, Teradata and Freshworks are named as major recruiters conducting campus interviews at CMI.
CMI's placement brochure describes its graduates - across all programmes - as going into software development, semiconductors, investment banking, analytics and healthcare, with campus placement having what the institute describes as an excellent track record for students seeking jobs through the process. On salary, the same brochure states that average pay packages have run ₹18-20 lakh per year in recent years institute-wide; CMI's own detailed table (see Placement Statistics) shows mean offers in the ₹13-21 LPA range and median offers in a similar band across the last several years, though again these are combined figures across CMI's programmes rather than a Data Science-only number.
Campus recruitment at CMI runs on a fixed annual calendar: companies make initial contact by August, CMI shares shortlisted student profiles by September, and the first round of interviews is typically completed by October, with additional rounds scheduled later in the year for students who miss the first cycle. Because MSc Data Science is one of the three cohorts CMI names as forming "the largest number of students participating in campus interviews," Data Science graduates are a core part of this recruiting pipeline each year, even though CMI has not published a separate salary breakdown isolating this specific degree.
CMI's own framing for this degree is explicitly industry-facing rather than academic-first: its programme materials describe the objective as preparing students for intensive data-analytics jobs in industry, in India and abroad, which is a different orientation from CMI's MSc Mathematics programme, which more commonly feeds into further research.
The most immediate, verifiable outcome channel is campus placement: CMI names MSc Data Science as one of the three groups making up the bulk of its campus-interview participants each year, with graduates recruited by firms such as Credit Suisse, Ernst & Young, TRDDC, Adobe, Zendrive, Teradata and Freshworks into analytics, data science and related technical roles across software, finance and analytics sectors.
Further academic study remains a live option even though it isn't the programme's primary design: nothing prevents an MSc Data Science graduate from pursuing a PhD (at CMI itself or elsewhere) in a data-science-adjacent field such as statistics, machine learning or applied mathematics, and CMI's faculty are themselves active researchers - for example, published work on Bayesian Gaussian process regression and portfolio risk analysis appears directly in the Bayesian Data Analysis elective's own reading list, giving interested students a concrete route into research collaboration during the degree itself.
CMI also runs Algolabs, an industry-facing society that has delivered training and project work for companies including Cognizant, Global Analytics, MRF and Tech Mahindra, giving another route by which graduates' skills connect to applied, ongoing industry work beyond the standard placement process. [NEEDS VERIFICATION: any officially published percentage of MSc Data Science graduates who pursue further postgraduate study rather than industry roles.]
[{"industry":"Banking and Finance","how_this_course_applies":"Applies statistical inference, predictive modelling, and risk analytics to solve strategic management and decision-making problems. The mandatory first-year industry internship provides direct exposure to real-world financial data challenges.","example_roles":["Risk Analyst","Data Scientist","Quantitative Researcher"]},{"industry":"Technology and Software","how_this_course_applies":"Leverages high-performance algorithms, distributed computing for big data, and machine learning to build scalable data pipelines and analytical tools. Students are explicitly trained to read, understand, and implement complex research papers for innovative technical solutions.","example_roles":["Data Engineer","AI/ML Engineer","Data Modeller"]},{"industry":"Analytics and Consulting","how_this_course_applies":"Utilizes data mining, advanced regression, and visualization techniques to derive actionable business insights. The curriculum's focus on theoretical rigour combined with hands-on practical expertise differentiates graduates in solving complex optimization and business problems.","example_roles":["Data Scientist","Business Analyst","Analytics Consultant"]},{"industry":"Research and Development","how_this_course_applies":"Builds on the institute's strong foundation in discrete mathematics, probability theory, and algorithmic design to drive innovation in areas like formal verification, large-scale data analysis, and computational modelling.","example_roles":["Research Scientist","R&D Engineer","Data Analyst"]}]
[{"name":"Simple Pseudocode","category":"Entrance Examination","used_for":"Testing algorithmic logic, variables, conditionals, and loops during the admission test without requiring prior knowledge of specific programming languages."},{"name":"Python","category":"Core Programming Language","used_for":"Programming and Data Structures coursework in Semester 1, forming the foundational skill set for subsequent machine learning and algorithm design courses."},{"name":"R","category":"Statistical Computing","used_for":"Probability and Statistics coursework in Semester 1, enabling rigorous statistical inference, modeling, and data analysis."},{"name":"SQL / RDBMS","category":"Database Management","used_for":"A dedicated 2-credit course in Semester 1 focused on relational database management systems and structured querying."},{"name":"Data Visualization Tools","category":"Data Presentation","used_for":"A dedicated 2-credit course in Semester 1 designed to teach effective communication of data insights and patterns."},{"name":"Machine Learning and Big Data Environments","category":"Advanced Applications","used_for":"Applied conceptually and practically in Semester 2 and 3 courses such as Data Mining, Machine Learning, Distributed Computing, Text Analytics, and Algorithmic Trading."}]
[{"stage":"Foundation and Concept Building","duration":"Mastery-based (Variable)","do_this":["Anchor your preparation strictly to the official 4-area entrance syllabus: School Level Mathematics, Discrete Mathematics, Probability Theory, and Programming.","Build core concepts in linear algebra, calculus, probability, and statistics before attempting complex or mixed problems.","Avoid passive reading; focus on understanding the 'why' behind mathematical properties and algorithmic logic."]},{"stage":"Topic-wise Application","duration":"Mastery-based (Variable)","do_this":["Solve focused, topic-wise questions immediately after studying a concept to identify exactly where your understanding breaks.","Practice reading and interpreting simple pseudocode (variables, conditionals, loops) to build comfort with the Programming section.","Ensure you are not studying chapters in isolation, but linking foundational concepts to their direct applications."]},{"stage":"PYQ Integration and Pattern Recognition","duration":"Mastery-based (Variable)","do_this":["Analyze previous year questions early to diagnose your baseline strengths and weaknesses.","Study how concepts are actually tested to develop depth, speed, and pattern recognition.","Use PYQ analysis to drive targeted gap-filling rather than random, unstructured practice."]},{"stage":"Mock Revision and Exam Strategy","duration":"Mastery-based (Variable)","do_this":["Take full-length mock tests simulating the exact 3.5-hour (2:00 PM to 5:30 PM) offline, pen-and-paper environment.","Train specifically to balance speed for Part A (20 objective questions) with rigorous, step-by-step reasoning for Part B descriptive components.","Revise actively by writing formulas, summarizing weak areas, and revisiting mistakes without looking at your notes."]}]
The MSc Data Science syllabus is built in two clearly sequenced layers: a fully core foundation across Semesters I-II, and a progressively more elective, specialisation-driven second year - with a compulsory industry internship acting as the hinge between them.
The practical implication for a new student is sequencing: Semester I's Mathematical Methods, Probability and Statistics with R, and Python programming are prerequisites in substance (even where not formally enforced) for Semester II's Data Mining and Machine Learning, Linear Algebra, and Distributed Computing and Big Data courses, so gaps left unresolved in the first semester tend to compound. Since the summer internship between years one and two is meant to shape which of the roughly fifteen available electives a student picks for Semesters III-IV, it is worth treating that internship not just as a requirement to complete but as a genuine input into planning a specialisation - whether that leans toward core ML and big-data infrastructure, quantitative finance, or applied domains like NLP or computer vision. Students preparing for the entrance exam should also note that the exam's own syllabus (school mathematics, discrete math, probability, and basic programming) is deliberately narrower than the full academic curriculum they will study once admitted.
There are two different things this question can mean, and CMI's public documentation only covers one of them clearly. For entrance-exam preparation, the closest equivalent to a "test series" is CMI's own archive of official past papers and solutions, published every year from 2018 through 2026 for MSc Data Science specifically - working through these chronologically, then attempting full timed 40-question, 3.5-hour mock papers under exam conditions, is the closest thing to a structured practice ladder that CMI itself provides.
For the actual academic programme once enrolled, CMI does not publish a standardised, publicly documented "unit test → midterm → final" structure across its roughly sixteen MSc Data Science courses; course-level assessment (assignments, mid-semester tests, end-semester examinations) is set by individual instructors rather than following one institute-wide public template. [NEEDS VERIFICATION: a detailed, course-by-course breakdown of internal unit-test and midterm weighting for MSc Data Science, which CMI has not published externally.]
CMI's MSc Data Science paper splits its 40 questions into two structurally different types, and understanding the split matters more than memorising a single "MCQ vs subjective" ratio.
Part A (20 questions, 40 of the 100 marks) is objective-type: mostly multiple-select questions where more than one option can be correct and full marks require selecting every correct option with no partial credit, alongside some "calculate and state the value" items requiring only a final numeric or symbolic answer with no explanation. Part B (20 questions, 60 of the 100 marks) is short-answer: each question needs a worked answer with a brief justification, and CMI explicitly allows partial credit here, rewarding correct method and reasoning even when the final answer is incomplete or slightly off. Based on CMI's own published sample papers and official solution keys, no negative marking is applied anywhere in this scheme - scores are purely additive, whether full, partial (Part B only), or zero.
This two-part structure is distinct from what many secondary sources describe for CMI's general BS-level paper (commonly reported as 10 objective questions worth 40 marks plus 6 descriptive questions worth 80 marks) - MSc Data Science instead uses a 20-plus-20, 40-marks-plus-60-marks split, a meaningfully different weighting between the objective and descriptive components. Topically, Part A questions have leaned toward probability, combinatorics, matrix algebra and short code-tracing puzzles in recent years, while Part B has included proof-style linear algebra questions, distribution-based probability derivations, and multi-step counting problems requiring full working.
Last updated: October 09, 2026
CMI does not publish a standalone placement breakdown specific to the MSc Data Science program. All published placement statistics are aggregated institute wide across all undergraduate and postgraduate programmes. However, MSc Data Science students, alongside MSc Computer Science and BS (Hons) Mathematics and Computer Science students, make up the largest share of participants in CMI campus interviews each year. This makes the institute wide median offer a highly reliable proxy for Data Science graduate outcomes.
Recent institute wide median offers have stabilized in the ₹16 LPA to ₹20 LPA range. Maximum offers exhibit high year to year volatility due to small cohort sizes, where a handful of exceptional offers can skew the top end without changing the typical outcome.
| Year | Maximum Offer | Mean Offer | Median Offer |
|---|---|---|---|
| 2024-25 | ₹37.5 LPA | ₹17.7 LPA | ₹16.0 LPA |
| 2023-24 | ₹25.8 LPA | ₹18.2 LPA | ₹18.6 LPA |
| 2022-23 | ₹47.0 LPA | ₹20.8 LPA | ₹20.7 LPA |
| 2021-22 | ₹62.0 LPA | ₹17.6 LPA | ₹16.0 LPA |
| 2020-21 | ₹18.4 LPA | ₹12.99 LPA | ₹13.5 LPA |
| 2019-20 | ₹20.0 LPA | ₹14.0 LPA | ₹13.35 LPA |
| 2018-19 | ₹16.54 LPA | ₹12.88 LPA | ₹14.8 LPA |
What most candidates get wrong here is anchoring their expectations to the maximum offer, such as the ₹62 LPA seen in 2021-22. The median offer is a significantly more reliable metric for evaluating typical CMI placements because it is immune to outlier spikes.
Your practical next action is to calculate your target preparation based on the median range of ₹16 LPA to ₹20 LPA, ensuring your financial planning remains realistic regardless of annual market fluctuations.
Historical placement records show MSc Data Science graduates securing specialized roles at top technology, finance, and analytics firms. The rigorous quantitative foundation of the program aligns directly with industry demands.
You can verify these historical trends and cohort details on the official CMI placement page.
Cracking the CMI Data Science entrance requires more than generic practice. MastersUp builds you a personalized, AI driven study plan that adapts to your real performance topic by topic.
Machine learning spots your weak and strong areas, surfacing curated practice questions exactly where you need them.
Learn on the go, one topic at a time, without distractions, ensuring deep conceptual retention.
Whether you have 12 months for 6 full revisions or just 3 months for a compact sprint, the platform adjusts your schedule. Even at 1 hour of daily practice, you see exactly where you stand against peers.
Because CMI never discloses a numeric cutoff or a percentage-based "safe score" for MSc Data Science, the honest way to think about safety margin here is different from asking what percentage to target, which is the framing that works for exams with published cutoffs.
What can be said with confidence, based on the exam's own structure, is where marks are easiest to protect and where they are easiest to lose. Part A's 40 marks come from 20 objective questions worth 2 marks each with no partial credit - CMI's own recent official solutions show many of these are "select all that apply" style questions, where missing even one correct option among several forfeits the full 2 marks for that question, so accuracy on multi-select items matters more than speed. Part B's 60 marks, by contrast, explicitly reward partial credit for correct method and reasoning, which means a student under time pressure is generally better off attempting more Part B questions with partial working shown than leaving several blank to double-check Part A.
Any specific numeric "safe score" (for instance, a claim like "aim for 60 out of 100") that circulates on forums or coaching pages is an informal community estimate at best, since CMI has never confirmed a threshold publicly and the paper's difficulty is not held constant year to year - the 2026 solutions, for example, include several multi-part probability and combinatorics questions noticeably more layered than the 2018 sample paper. [NEEDS VERIFICATION: any specific numeric safe-score benchmark for MSc Data Science - none could be confirmed against an official CMI source.]
CMI's rigorous 64-credit MSc Data Science programme for quantitative graduates, featuring a dedicated 3.5-hour entrance exam with zero negative marking.
Last updated: October 08, 2026
The CMI MSc Data Science entrance exam is a dedicated 3.5-hour, 100-mark pen-and-paper test held annually, operating completely independently from the Chennai Mathematical Institute's Mathematics or Computer Science papers.
Running as its own distinct question paper since the programme's first intake in 2018, this national-level offline examination is conducted at roughly 37 exam centres across India. A defining feature is its unique afternoon timing: the paper runs from 2:00 PM to 5:30 PM, granting candidates an extra 30 minutes compared to the standard 3-hour window given to other CMI postgraduate tests to accommodate its larger descriptive component.
The exam consists of 40 questions divided into two parts totaling 100 marks, with zero negative marking and partial credit awarded for descriptive reasoning.
| Section | Question Type | Questions | Marks per Question | Total Marks |
|---|---|---|---|---|
| Part A | Objective | 20 | 2 | 40 |
| Part B | Short Answer Descriptive | 20 | 3 | 60 |
What most candidates get wrong here: Many assume standard multiple choice negative marking applies to Part A. It does not. Failing to attempt a question because of a perceived penalty is a critical strategic error. You must attempt every objective question and show your working clearly in Part B to harvest partial marks.
Your practical next action today is to visit the official CMI admissions syllabus page, download the recent question papers, and time yourself solving only the Matrices and Conditional Probability sections, as these specific chapters consistently yield the highest question volume.
You are eligible if you hold or are in the final year of a B.A., B.Sc., B.Math., B.Stat., B.E., or B.Tech. degree with a background in Mathematics, Statistics, or Computer Science.
This wording is deliberately broad. It explicitly admits engineers, physicists, and other quantitative graduates, dispelling the common myth that only pure statistics or computer science majors can apply. Final year undergraduates who expect to complete their degree by the start of the relevant academic session are fully permitted to apply provisionally.
The entrance syllabus focuses on school level mathematics, discrete mathematics, probability, and programming, which is entirely distinct from the programme syllabus taught after enrollment.
Based on an analysis of 293 past year questions, the historical weightage dictates your preparation priority:
CMI Data Science Entrance Exam: Pattern, Syllabus & PYQs
Master the CMI MSc Data Science entrance exam. Get the exact 3.5-hour pattern, no-negative-marking rule, eligibility for B.Tech/B.Sc, and PYQ weightage.
Last updated: October 08, 2026
A successful CMI MSc Data Science preparation strategy requires a structured 6-month plan that balances objective speed with descriptive mathematical reasoning. Based on 293 past year questions, you must prioritize School Level Mathematics and Probability Theory while dedicating specific weekly hours to writing out proofs for Part B partial credit.
A realistic 6-month timeline divides preparation into a 3-month foundation phase for topic-wise mastery and a 3-month consolidation phase focused on full-length mock tests and descriptive writing.
Dedicate 1.5 to 2 hours daily to conceptual building. Focus heavily on School Level Mathematics and Discrete Mathematics. Solve foundational problems systematically rather than rushing through advanced topics.
Scale your daily study time to 2 to 3 hours. Shift the focus from learning new concepts to applying them under timed conditions and refining your descriptive writing.
Strategic preparation demands prioritizing chapters that historically dominate the 293 analyzed past year questions, ensuring maximum return on your study time.
| Chapter | Questions | Weightage |
|---|---|---|
| Matrices and Linear Systems | 13 | 4.44% |
| Divisibility, GCD, LCM and Modular Arithmetic | 12 | 4.10% |
| Conditional Probability, Bayes' Theorem and Independence | 12 | 4.10% |
| Determinants, Matrix Inverses and Special Matrices | 11 | 3.75% |
The most authoritative resources for the quantitative sections align directly with the official CMI entrance syllabus, avoiding generic material that misses the descriptive depth required.
1
Test of Mathematics at the 10+2 Level (TOMATO): The definitive core text for School Level Mathematics and Discrete Math problem solving.
2
Objective Mathematics by RD Sharma: Excellent for foundational drills and building speed in algebraic manipulations.
3
Introduction to Probability by Sheldon Ross: Highly recommended for building the conceptual clarity needed for the 22.87 percent Probability Theory weightage.
4
CMI Data Science Preparation Strategy: 6-Month Plan Master the CMI MSc Data Science entrance exam with our 6-month preparation strategy, high-yield PYQ triage, and the best books for Part A and Part B success. Last updated: October 08, 2026 The CMI MSc Data Science admission procedure is a strict chronological funnel starting with a March application, a May entrance exam, and results in June. Unlike CMI Mathematics, interviews are strictly discretionary, and the institute does not publish a public rank list. The entire admission cycle runs over a focused five-month window. You must complete each sequential milestone to secure your seat in the August intake. Interviews for the MSc Data Science programme are strictly at the discretion of the Admissions Committee and are not mandatory for all shortlisted candidates. Aspirants often assume every shortlisted candidate faces a mandatory panel interview like the PhD or MSc Mathematics applicants. That is incorrect. The committee typically only calls candidates to clarify borderline academic records or specific prior coursework. If you clear the exam with a strong quantitative background, you may receive direct admission without facing an interview panel. CMI does not officially publish a formal rank versus marks cutoff list or a public merit list for the Data Science programme. Admission outcomes are heavily dependent on the overall candidate performance distribution and the specific difficulty of that year's paper. Instead of hunting for a rigid safe score on third-party forums, your practical next action today is to review your Part B descriptive writing. Maximizing partial credit through clear mathematical proofs is the most reliable way to push your total score above the unspoken selection threshold. The standard annual chronology follows a predictable spring and summer schedule. Track the key milestones below to plan your academic calendar. Navigating an opaque admission process requires absolute clarity on where you stand before the exam. MastersUp builds a personalized, AI-driven study plan that adapts to your real performance. 1 CMI MSc Data Science Admission Procedure & Timeline Understand the CMI MSc Data Science admission procedure, discretionary interview policy, cutoff transparency, and the official step-by-step timeline. Last updated: October 08, 2026 The CMI MSc Data Science curriculum is a rigorous 64-credit, four-semester programme designed to transition quantitative graduates into advanced predictive analytics and machine learning roles. As per the official institute notification, regular full-semester courses carry 4 credits each, and you must complete a minimum of 16 regular courses to earn the degree. This structure is entirely distinct from the entrance exam syllabus, focusing instead on applied mathematics, statistical computing, and algorithmic design. The programme builds foundational skills in the first year before advancing to specialized machine learning and distributed systems. What most candidates get wrong here is assuming the degree continues to test basic discrete mathematics. The actual coursework rapidly shifts to applied statistical modeling and programming. Your practical next action is to visit the official CMI course structure page and map your undergraduate background to these core requirements. Identify any foundational gaps in Python or linear algebra now. Review Entrance vs Degree Syllabus A mandatory summer internship is scheduled between Semester II and Semester III. This component bridges the gap between theoretical machine learning concepts learned in the first year and real-world predictive data analysis applications. While the exact grading rubric varies, completing this internship is a strict requirement for degree progression. Semesters III and IV allow you to tailor your degree through specialized electives. This flexibility enables you to build a versatile profile for specific industry roles or academic research. An Industry Project is also available as an elective, allowing you to substitute a traditional classroom course with applied, supervised research. CMI MSc Data Science Curriculum: Semesters & Electives Explore the official CMI MSc Data Science curriculum: 4-semester structure, 64-credit requirement, core subjects, compulsory internship, and advanced electives. Last updated: October 08, 2026 A typical day for a CMI MSc Data Science student revolves around 2 to 3 lectures daily, each lasting approximately 75 minutes, balanced with intense self-study. You will navigate a rigorous academic schedule while preparing for an internship drive that begins as early as mid-October of your first semester. As per the official institute notification, on-campus hostel accommodation is explicitly not guaranteed for MSc Data Science students, unlike other CMI programmes. You must be prepared to explore alternative housing. Most students secure paying guest accommodations or rent flats in groups within the nearby SIPCOT, Navallur, or Padur areas. Factor this logistical step into your pre-semester planning. Verify official CMI hostel policies What most candidates get wrong here is assuming the first year is purely academic and relaxed. Internship hiring for Data Science students begins as early as mid-October of the first semester. This requires you to balance foundational coursework with immediate project building and interview preparation. A compulsory 2 to 3 month summer internship between your first and second year serves as the primary pipeline for pre-placement offers. Your practical next action is to build a foundational coding or machine learning project before the semester begins. This gives you concrete material to discuss during those critical early October interviews. The CMI campus is located inside the SIPCOT IT Park, approximately 30 km outside the Chennai city center. While this offers a peaceful environment for studying, it requires planning for daily needs. The institute provides regular transportation arrangements for students to visit the city. Shared autos are readily available from the main gate. For food, many students supplement mess meals with local tiffin services. Weekends are frequently spent on short trips to nearby destinations like Mahabalipuram or Pondicherry. Preparing for the CMI entrance requires more than generic practice. MastersUp builds each learner a personalized, AI-driven study plan that adapts to your real performance. Machine learning tracks your topic-by-topic performance, spots weak areas, and adjusts what you practice next. Cocoon is our focus-mode flow for learning on the go, serving one curated topic at a time without distractions. CMI MSc Data Science Student Life: Daily Routine & Housing Discover the daily routine of a CMI MSc Data Science student. Learn about the lecture schedule, early internship prep, and housing realities near SIPCOT. Last updated: October 08, 2026 The campus life for CMI MSc Data Science students is defined by a tight-knit, highly intellectual peer culture set within the peaceful environment of the SIPCOT IT Park. While academic rigor is high, you will find a vibrant balance through student-run initiatives like the annual Tessellate fest, active coding and movie clubs, and a collaborative atmosphere among diverse quantitative graduates. The peer culture at CMI is intimate and highly collaborative, bringing together students from diverse quantitative backgrounds like engineering, mathematics, and statistics. The small campus size naturally fosters strong mentorship and peer learning. You will frequently share open discussion areas and project work with peers who are national olympiad medalists or experienced industry professionals, creating an environment where academic support is always within reach. Extracurricular life at CMI centers around active student societies and the annual Tessellate fest, which provides a creative reprieve from rigorous coursework. This student-run event features chess tournaments, cubing competitions, quizzes, debates, and the S.T.E.M.S. competition. Beyond the annual fest, regular club activities include weekend movie screenings in the seminar hall, coding club hackathons, literature society discussions, and intramural sports tournaments. As per the official CMI MSc Programme notification, on-campus hostel accommodation is explicitly not guaranteed for the MSc Data Science programme. What most candidates get wrong here is assuming they will automatically secure a campus room like students in other CMI degree programmes. You must proactively explore practical alternatives. Most students successfully secure unisex paying guest accommodations or rent flats in groups within the nearby SIPCOT, Navallur, or Padur areas. Your practical next action is to join CMI student community groups a month before the semester starts to coordinate flat hunting and roommate matching with incoming batchmates. Verify official CMI hostel policies Managing daily life requires planning, as the CMI campus is located approximately 30 km outside the Chennai city center. This isolation provides a highly focused and peaceful environment for studying. The institute provides regular shuttle services for city commutes, and shared autos are readily available from the main gate. Students frequently balance campus mess food with local tiffin services and plan weekend escapes to nearby destinations like Mahabalipuram or Pondicherry to recharge. Preparing for the CMI entrance requires more than generic practice. MastersUp builds each learner a personalized, AI-driven study plan that adapts to your real performance. Machine learning tracks your topic-by-topic performance, spots weak areas, and adjusts what you practice next. Cocoon is our focus-mode flow for learning on the go, serving one curated topic at a time without distractions. Revision depth tracking ensures you hit the optimal number of revisions, from 6 full cycles over 12 months down to a compact sprint mode when time is short. Even at 1 hour of daily practice, you can see exactly where you stand, topic by topic, against other students on the platform. The admission cycle for the MSc Data Science programme follows a strict chronological funnel. Exact dates vary slightly each year, but the sequence remains consistent.Preparation Strategy Seo Title
Preparation Strategy Seo Description
Admission Procedure Blog
CMI MSc Data Science Admission Procedure Overview
Step-by-Step Admission Procedure
Interview Policy: Mandatory or Discretionary?
What most candidates get wrong here:
Cutoff and Merit List Transparency
Official Admission Timeline
Milestone Expected Schedule Application Window Admit Card Release Entrance Examination Result Declaration Academic Session Start The MastersUp Advantage for CMI Aspirants
Admission Procedure Seo Title
Admission Procedure Seo Description
Curriculum Blog
CMI MSc Data Science Curriculum Overview
Semester-Wise Core Course Breakdown
Semester Core Courses Credit Structure Semester I Mathematical Methods (Analysis), Probability and Statistics with R, Programming and Data Structures with Python, Visualization, RDBMS and SQL 4 credits each, except Visualization and RDBMS/SQL which are 2 credits each Semester II Linear Algebra and its Applications, Data Mining and Machine Learning, Algorithms, Distributed Computing and Big Data 4 credits per course Semester III Regression and Classification, Advanced Machine Learning, plus two student-selected electives 4 credits per course Semester IV Four student-selected electives 4 credits per course (or accumulated 1 to 2 credit short electives) Compulsory Summer Internship
Elective Specializations and Advanced Topics
Quantitative Finance
Advanced Analytics
Systems and Modeling
Curriculum Seo Title
Curriculum Seo Description
Day In Life Blog
A Realistic Daily Schedule for CMI Data Science Students
Time Activity Details 9:00 AM Morning Routine Breakfast at the campus mess, followed by a short walk to the academic block. 9:30 AM Core Lectures 75-minute sessions covering foundational topics like Probability with R or Data Structures in Python. 1:00 PM Lunch Break Mess lunch or exploring local tiffin services near the SIPCOT gate. 2:30 PM Self-Study and Projects Library or open discussion areas for machine learning assignments and group project work. 5:00 PM Extracurriculars Data Science colloquiums, coding club meetings, or casual evening walks. The Housing Reality: Navigating Accommodation
Balancing Coursework and Early Career Preparation
Campus Logistics: Commute, Food, and Weekends
The MastersUp Advantage: Precision Over Guesswork
Day In Life Seo Title
Day In Life Seo Description
Campus Life Blog
CMI MSc Data Science Campus Life Overview
The CMI Campus Culture and Peer Environment
Student Clubs and the Annual Tessellate Fest
Hostel Realities and Off-Campus Living
Work-Life Balance and Campus Logistics
The MastersUp Advantage: Precision Over Guesswork
Official Academic Timeline