IIT JAM Optimization Techniques for Statisticians: practice questions
IIT JAM - Mathematical Statistics (MS) · India · IITs (JAM 2026 organised by IIT Bombay)
Original practice questions written to the published syllabus · not past papers
10worked here, free
39written to this exam in this subject
10syllabus units
About IIT JAM
IIT JAM - Mathematical Statistics (MS) is a postgraduate admission exam run by IITs (JAM 2026 organised by IIT Bombay), for candidates in India. Computer-based, 3 hours, 60 questions, 100 marks, English. Section A 30 MCQs (10 x 1 mark, 20 x 2 marks); Section B 10 MSQs (2 marks, no negative or partial marking); Section C 20 NAT (10 x 1, 10 x 2, no negative). MCQ wrong answers lose 1/3 or 2/3 mark.
Checked against the exam body's published information. Formats and fees change; confirm the current edition there before you book.
Optimization Techniques for Statisticians in IIT JAM
The optimisation theory statisticians use: linear programming with simplex and duality, transportation and assignment, integer and goal programming for farm and crop planning, non-linear optimisation with Kuhn-Tucker conditions, and the numerical optimisers inside likelihood estimation. The Agribusiness track teaches operations research for managers; this goes to the theory.
Our 39 IIT JAM questions in this subject: 3 easy, 29 medium, 7 hard, across 10 units.
10 practice questions, worked
Choose an answer, then open the working. Each explanation says why the right answer is right and why the others are not.
These 10 were written to IIT JAM and are not shown on any other exam's page.
Question 1 · medium
If the primal LP is unbounded, what is the dual?
Infeasible
Degenerate
Optimal with value zero
Unbounded
Answer and working
A. Infeasible
By weak duality, an unbounded primal leaves no feasible dual.
Question 2 · medium
In the Hungarian method, when has an optimal assignment been reached?
When one line covers all zeros
After exactly n iterations
When all entries are zero
When n lines cover the zeros
Answer and working
D. When n lines cover the zeros
n covering lines guarantee n independent zeros for a complete assignment.
Question 3 · medium
Which class of problems is NP-hard in general?
The balanced transportation problem
Linear programming
Integer programming
Unconstrained convex quadratic minimisation
Answer and working
C. Integer programming
LP and transportation are polynomially solvable; general IP is NP-hard.
Question 4 · medium
In what sense is goal programming 'satisficing'?
It targets aspiration levels
It always finds the global optimum
It ignores constraints
It maximises a single profit
Answer and working
A. It targets aspiration levels
Targets are set by the decision maker, and the solution minimises distance to them.
Question 5 · medium
Why should a goal-programming farm plan be re-solved under several weightings?
Goal programmes have no unique feasible region
Solvers need several runs to converge
Weights do not affect the solution
Results depend strongly on weights and targets
Answer and working
D. Results depend strongly on weights and targets
Sensitivity to weights shows the trade-offs to the decision maker instead of hiding them.
Question 6 · medium
In maximising xy subject to x + y = 10, λ = 5. By about how much does the optimum rise if the constraint becomes x + y = 10.1?
0.05
5
0.1
0.5
Answer and working
D. 0.5
λ is the marginal value of the constraint: Δ ≈ λ × 0.1 = 0.5 (exactly 5.05² − 25 = 0.5025).
Question 7 · medium
For normally distributed income with mean absolute deviation 10, approximately what is the standard deviation?
15.7
8.0
10.0
12.5
Answer and working
D. 12.5
MAD = σ√(2/π), so σ = MAD × √(π/2) = 12.5. Multiplying by √(2/π) goes the wrong way.
Question 8 · medium
In which direction are deterministic DP recursions usually solved?
In random order
From the middle stage
Backwards
Forwards only
Answer and working
C. Backwards
fₙ depends on fₙ₊₁, so the last stage is solved first (forward formulations also exist).
Question 9 · medium
What is the main weakness of steepest-descent (gradient) methods?
Slow zig-zagging
They converge in one step
They cannot handle smooth functions
They need second derivatives
Answer and working
A. Slow zig-zagging
With correlated parameters the contours are elongated and gradient steps oscillate.
Question 10 · hard
Why is reparameterising λ as η = log λ helpful for the Poisson likelihood 30 log λ − 10λ?
It is concave in η
It makes the MLE zero
It doubles the convergence rate exactly
It removes the need to optimise
Answer and working
A. It is concave in η
30η − 10e^η has second derivative −10e^η < 0, so Newton steps are well behaved and λ stays positive.
A study plan for Optimization Techniques for Statisticians in IIT JAM
This plan follows where our questions actually fall, so the time goes where the marks are likely to be. It is our reading of the syllabus, not the exam body's weighting.
First, the heavy units (69% of the 39 questions): Numerical optimisation: gradient, Newton and quasi-Newton methods (9, 23%); Non-linear programming: Lagrange multipliers and Kuhn-Tucker conditions (8, 21%); Dynamic programming (5, 13%); Integer programming: branch and bound (5, 13%).
Read the manual chapters for these first and do every worked question you can find on them.
Then the middle band: Goal programming for multi-objective farm planning (3, 8%); Duality and sensitivity analysis (2, 5%); Linear programming: formulation and graphical solution (2, 5%); Quadratic programming and risk programming (MOTAD) (2, 5%).
Aim for one sitting per unit, then mixed questions across them.
Then the lighter units: Simplex method, Big-M and two-phase methods (2, 5%); Transportation and assignment problems (1, 3%).
Learn the definitions, classifications and standard numbers; these units are more often recalled than worked.
Leave time for the hard ones: Duality and sensitivity analysis (1 hard of 2), Numerical optimisation: gradient, Newton and quasi-Newton methods (3 hard of 9), Dynamic programming (1 hard of 5).
Come back to these after a first pass, untimed, and read every explanation, including for the options you rejected.
Bank the easy marks: Dynamic programming (1 easy of 5); Non-linear programming: Lagrange multipliers and Kuhn-Tucker conditions (1 easy of 8).
Do these early in revision so they are fast on the day.
Finish timed. Sit the subject as a timed set without notes, mark it, and send each wrong answer back to its unit above.