Quick answer. Section 12 of the IIT JAM Mathematical Statistics syllabus is Stochastic Processes, and it is tightly bounded. It covers discrete time Markov chains and the Poisson process, and nothing beyond them. Continuous time Markov chains, birth and death processes, Brownian motion, queueing theory and renewal theory are all outside the syllabus. Knowing where the boundary sits saves more time here than anything else, because the standard textbooks all continue well past it.
Why This Section Catches People Out
Two reasons, and neither is difficulty.
Most undergraduate Statistics programmes treat stochastic processes as an elective, or cover it in a single hurried chapter. So a large share of JAM MS candidates meet this material for the first time while preparing.
The books all go far past the syllabus. Open Medhi or Ross and you will find continuous time chains, birth and death processes, Brownian motion, queueing and renewal theory sitting immediately after the chapters you actually need. Nothing tells you to stop.
The result is candidates who study three times more than the section asks, and still miss the parts it does ask.
Exactly What Is In It
The official syllabus is short. Here it is, in full, split into the two things it covers.
Discrete time Markov chains
- Transition probability matrix
- Higher order transition probabilities
- Markov chain as a graph
- Chapman Kolmogorov equation
- Classification of states and chains
- Stability of a Markov chain, meaning stationary and limiting distributions
Poisson process
- The Poisson process and its properties
- Interarrival and waiting times
That is the entire section. Two topics, roughly ten sub topics.
The full JAM MS syllabus, all 12 sections
Exactly What Is Not In It
This list is worth more than the one above, because it is where the wasted weeks go.
| Not in the JAM MS syllabus | Where you will find it in the books |
|---|---|
| Continuous time Markov chains | Immediately after discrete time chains in every text |
| Birth and death processes | Usually the next chapter |
| Brownian motion | Later chapter |
| Queueing theory, M/M/1 and similar | Often the longest chapter in the book |
| Renewal theory | Later chapter |
| Martingales | Advanced chapters |
Stop when the book moves past the Poisson process.
Worth knowing for comparison: GATE ST does include birth and death processes and Brownian motion. UGC NET Statistics includes the gambler’s ruin problem, random walks and M/M/1 queues. RBI DSIM includes Brownian motion and martingales.
JAM MS includes none of those. If you are preparing more than one of these examinations, keep the boundaries separate or you will over prepare for JAM and think you are being thorough.
How Deep to Go
The JAM MS paper is 60 questions across twelve sections. Section 12 is one of twelve, and it is one of the shorter ones.
That gives you a sensible calibration:
| Depth | Right for Section 12? |
|---|---|
| Recognise a transition matrix and read it | Yes, essential |
| Compute n step transition probabilities | Yes |
| Classify states as transient, recurrent, periodic, absorbing | Yes |
| Find a stationary distribution for a small chain | Yes |
| Apply Chapman Kolmogorov | Yes |
| Poisson process properties, interarrival and waiting time distributions | Yes |
| Prove limit theorems for Markov chains rigorously | No. Understand the statement and when it applies. |
| Continuous time or queueing extensions | No |
Aim for confident computation and clear understanding, not for proofs. JAM is an objective paper with MCQ, MSQ and NAT questions. It asks whether you can work with a chain, not whether you can prove ergodicity.
The Five Things Actually Worth Drilling
If your time on this section is limited, spend it here.
1. Reading a transition probability matrix. Rows sum to one. Row i column j is the probability of moving from state i to state j. Half the errors in this topic come from reading the matrix transposed.
2. Higher order transitions. The n step transition matrix is the matrix raised to the power n. Practise this by hand on 2×2 and 3×3 chains until it is automatic.
3. Classification of states. Transient against recurrent, periodic against aperiodic, absorbing. This is where MSQ questions live, because a state can carry more than one property and MSQ has no partial marking.
4. Stationary distributions. Solve pi times P equals pi with the components summing to one. On a small chain this is a short linear system, and it appears often.
5. Poisson process interarrival times. Interarrival times are exponential and independent. The waiting time to the nth arrival is Gamma. The link between the Poisson process and the exponential and Gamma distributions is the most examinable idea in this half of the section, and it also connects back to Section 5 on univariate distributions.
Where to Read It
Sheldon Ross, Introduction to Probability Models. The Markov chains chapter and the chapter on the exponential distribution and the Poisson process. Stop there.
J. Medhi, Stochastic Processes, if Ross feels too brief. Same two areas, more detail.
Either one, not both. And in both cases, stop at the Poisson process. The rest of the book is a different syllabus.
The full chapter level book guide
The Previous Year Paper Problem
Section 12 stands as a named section in the current syllabus. Older JAM MS papers were set under a syllabus that presented the sections differently, so previous year papers will not give you a reliable count of how many questions to expect here.
Two consequences.
Do not judge its importance by how often it has appeared. Judge it by the fact that it is one of twelve named sections.
Do not skip it because past papers looked light on it. A named section is a section the paper can draw from.
How Long This Should Take
Ten to fourteen days, if you are meeting the material for the first time and studying it properly.
| Days | What |
|---|---|
| 1 to 3 | Markov chain basics: states, transition matrix, the chain as a graph |
| 4 to 6 | Higher order transitions and Chapman Kolmogorov, worked by hand |
| 7 to 9 | Classification of states and chains |
| 10 to 11 | Stationary and limiting distributions |
| 12 to 14 | Poisson process, interarrival and waiting times |
Do not leave it until January. It is short enough to feel postponable and unfamiliar enough that postponing it means never doing it.
Frequently Asked Questions
What is in Section 12 of the JAM MS syllabus?
Discrete time Markov chains, covering the transition probability matrix, higher order transition probabilities, the chain as a graph, the Chapman Kolmogorov equation, classification of states and chains, and stationary and limiting distributions. Plus the Poisson process with interarrival and waiting times.
Is Brownian motion in the JAM MS syllabus?
No. Brownian motion, continuous time Markov chains, birth and death processes, queueing theory and renewal theory are all outside the JAM MS syllabus, although some of them appear in GATE ST, UGC NET Statistics or RBI DSIM.
Are queueing models in JAM MS?
No. M/M/1 and similar queueing models are not in the JAM MS syllabus. They appear in UGC NET Statistics.
How much time does Section 12 need?
Around ten to fourteen days if the material is new to you. It is one of the shorter sections of the twelve.
Which book covers stochastic processes for JAM MS?
Sheldon Ross’s Introduction to Probability Models, reading only the Markov chains chapter and the chapter on the exponential distribution and the Poisson process. J. Medhi’s Stochastic Processes is the alternative if you want more detail.
How many questions come from Section 12 in JAM MS?
The paper does not publish a section wise distribution, and older papers were set under a differently presented syllabus, so past frequency is not a reliable guide. Treat it as one of twelve named sections.
Do I need to prove Markov chain limit theorems?
No. JAM is an objective paper. Understand the statements and the conditions under which they apply, and be able to compute confidently on small chains.
