Exam guide & study plan
How the GATE Data Science and Artificial Intelligence (DA) paper is structured and marked, how its marks have been spread across subjects, how to attack each question type, and a 16-week plan for working through this guide.
The paper at a glance
| Item | Detail |
|---|---|
| Paper code | DA — Data Science and Artificial Intelligence |
| Mode | Computer-based test, with an on-screen virtual scientific calculator |
| Duration | 3 hours |
| Questions | 65 |
| Total marks | 100 |
| General Aptitude (GA) | 10 questions — 5 × 1 mark + 5 × 2 marks = 15 marks |
| Subject (DA) | 55 questions — 25 × 1 mark + 30 × 2 marks = 85 marks |
Question types and marking
| Type | What you do | Correct | Wrong | Partial credit |
|---|---|---|---|---|
| MCQ — multiple choice | pick the one correct option of four | +1 or +2 | −1/3 (1-mark) or −2/3 (2-mark) | — |
| MSQ — multiple select | pick all correct options (one or more) | +1 or +2 | 0 | none — every correct option and no wrong one |
| NAT — numerical answer | type a number (a range of answers is accepted) | +1 or +2 | 0 | — |
Where the marks have come from
GATE DA was first held in 2024. The official papers do not publish a subject-wise breakdown; the figures below are approximate counts from third-party paper analyses (GeeksforGeeks, MADE EASY, PhysicsWallah), rounded to whole marks. Use them to prioritise, not as a promise.
| Subject | 2024 (≈ marks) | 2025 (≈ marks) | This guide |
|---|---|---|---|
| Probability and Statistics | 16 | 18 | Probability & Statistics |
| Linear Algebra | 10 | 17 | Linear Algebra |
| Calculus and Optimization | 8 | 8 | Calculus |
| Programming, DS and Algorithms | 21 | 16 | Programming & DSA |
| DBMS and Warehousing | ≈ 8† | ≈ 9† | DBMS |
| Machine Learning | 11 | 14 | Machine Learning |
| Artificial Intelligence | 11 | 3 | AI |
| General Aptitude | 15 | 15 | General Aptitude |
† By difference: the remainder of the 85 subject marks after the other six subjects.
Two patterns stand out. The mathematics (probability, linear algebra, calculus) is about 40% of the paper — more than any single CS subject. And programming and algorithms questions are mostly about tracing code and data structures, not about writing code.
Attacking the paper
- First pass (≈ 90 min): answer everything you can solve quickly and confidently, in any order. Mark the rest for review.
- Second pass (≈ 60 min): the longer 2-mark NATs and MSQs you marked.
- Last 20–30 min: fill every unanswered NAT and MSQ with your best attempt; decide on MCQ guesses by elimination.
MSQ discipline: evaluate each option independently, as a true/false question. MSQs often have 2 or 3 correct options; a single correct option is also possible.
A 16-week study plan
The pages are ordered so that later topics build on earlier ones. Mark each topic studied as you go — the home page tracks your progress, and the question bank tracks which questions you got right.
| Weeks | Focus | Pages |
|---|---|---|
| 1–3 | Probability and statistics | all 9 pages; do every practice question |
| 4–6 | Linear algebra | all 8 pages; eigenvalues and SVD twice |
| 7 | Calculus and optimisation | 5 pages |
| 8–9 | Programming, data structures, algorithms | 10 pages; trace code by hand |
| 10 | DBMS and warehousing | 8 pages; write the SQL queries yourself |
| 11–13 | Machine learning | 13 pages — builds on probability and linear algebra |
| 14 | Artificial intelligence | 8 pages |
| ongoing | General aptitude | 3 pages; 20 minutes a day of mixed GA questions |
| 15–16 | Revision and mocks | formula sheet, wrong questions in the question bank, one full mock test every few days |
Using this guide
- Each topic page explains the theory with worked examples, then ends with practice questions in GATE format (MCQ / MSQ / NAT, 1 or 2 marks) and full solutions.
- The question bank collects every question from every page, with filters by section, topic, type, marks and status.
- The mock test builds a timed paper on the GATE pattern from the bank, with negative marking and a score report.
- The formula sheet is a one-page-per-subject summary for the last weeks.
Progress is saved in your browser only.
Sources
- Official syllabus: GATE 2027 DA syllabus (gate2027.iitm.ac.in).
- Pattern and marking scheme: official GATE information brochures.
- Subject-wise weightage (approximate, third-party): GeeksforGeeks "GATE DA 2025 paper analysis", MADE EASY blog "GATE DA subject-wise weightage", PhysicsWallah "GATE DA exam pattern".