GATE DA syllabus and paper pattern

How the DA paper is laid out

Questions carry either 1 or 2 marks. MCQs have a single right option; MSQs can have more than one, and you must choose all the correct ones with no extra pick to score. NAT answers are typed with a virtual keypad. Only wrong MCQ answers lose marks.

The general aptitude part is identical across GATE papers and is covered on the GATE aptitude syllabus page. For the full list of papers and common rules, start from the GATE syllabus overview.

GATE DA syllabus: the mathematics sections

Probability and statistics

• Counting, axioms of probability, independent and mutually exclusive events
• Marginal, conditional and joint probability, Bayes theorem, conditional expectation and variance
• Mean, median, mode, standard deviation, correlation and covariance
• Random variables with PMF, PDF and CDF, and the common discrete and continuous distributions, including normal, t and chi squared
• Central limit theorem, confidence intervals, and z, t and chi squared tests

Linear algebra

• Vector spaces, subspaces, linear dependence and independence
• Projection, orthogonal, idempotent and partitioned matrices, quadratic forms
• Systems of linear equations and Gaussian elimination
• Eigenvalues and eigenvectors, determinant, rank, nullity, LU decomposition and singular value decomposition

Calculus and optimisation

• Functions of a single variable: limit, continuity, differentiability
• Taylor series, maxima and minima, optimisation in one variable

Computing and data sections

Programming, data structures and algorithms

• Programming in Python
• Stacks, queues, linked lists, trees and hash tables
• Linear and binary search; selection, bubble and insertion sort; divide and conquer with mergesort and quicksort
• Graph theory basics, graph traversals and shortest paths

Database management and warehousing

• ER model, relational model, relational algebra and tuple calculus
• SQL, integrity constraints and normal forms
• File organisation and indexing
• Data types and transformation: normalisation, discretisation, sampling and compression
• Warehouse modelling: schemas for multidimensional data, concept hierarchies and measures

Machine learning and the reasoning section

Machine learning

• Supervised learning: regression and classification; simple, multiple and ridge regression; logistic regression; k nearest neighbours; naive Bayes; linear discriminant analysis; support vector machines; decision trees
• Model evaluation: bias and variance trade off, leave one out and k fold cross validation
• Neural networks: multi layer perceptron and feed forward networks
• Unsupervised learning: k means and k medoids clustering, hierarchical clustering with single and multiple linkage, and dimensionality reduction through principal component analysis

Search, logic and uncertainty

• Uninformed, informed and adversarial search
• Propositional and predicate logic
• Reasoning under uncertainty: conditional independence, exact inference through variable elimination, and approximate inference through sampling

Where the marks tend to come from

Because probability, statistics, linear algebra and calculus form three of the seven sections, and machine learning itself leans on the same maths, a strong mathematics base lifts your score across the paper. Treat probability and linear algebra as the core of your preparation rather than as background reading.

Programming questions are usually about tracing Python code or reasoning about an algorithm's output and cost, not writing long programs. Database questions favour SQL queries, relational algebra and normal forms. The reasoning section is short but precise, so a few hours on search algorithms and inference rules gives good returns.

Five months, section by section

1. Month 1: probability, statistics and linear algebra with plenty of numerical practice.
2. Month 2: calculus and optimisation, then Python, data structures and algorithms.
3. Month 3: databases and warehousing, and the search, logic and uncertainty section.
4. Month 4: machine learning, linking each model back to the maths behind it.
5. Month 5: general aptitude, full length mock tests on the official virtual calculator, and past DA papers.

Past papers for DA exist only for a few cycles so far, so add questions from the CS paper's algorithms and database sections and from the statistics paper for extra practice.

Books and resources

Use standard undergraduate texts in probability and statistics, linear algebra and calculus, a well known algorithms textbook, a database systems textbook, and an introductory machine learning text. Past DA question papers with official keys are posted on the GATE portal of each organising institute. If you also plan to sit the computer science paper, the GATE CSE syllabus shows how much overlaps.