Create an account to study the free first module. Start free.
The ZC418 handout runs 16 contact sessions across 6 modules — from a first look at a dataset through Bayes, inference, regression, and time series, to Gaussian mixtures and EM. All 16 course sessions are built and open below, so the course lines up with the handout you will be examined against. Every formula carries the textbook section and printed page it came from, so you can check it in Devore (Source) or Brockwell & Davis (Source).
Statistical methods is the layer between a dataset and a claim. Almost every machine-learning result you will ever defend rests on something taught here: a distribution assumed normal, a sample treated as representative, a p-value read as a conclusion. The course follows the official AIML ZC418 handout contact session by contact session — all sixteen of them, including the two revision modules — so the page you read in week 6 is the page your exam is set from.
It begins where statistics actually begins: describing data you can see, then the axioms that make probability a number rather than a feeling. From there it moves through conditional probability and Bayes, random variables and the five distributions you will use for the rest of your life, then inference — sampling, the central limit theorem, interval estimation, and hypothesis testing. The last third is prediction: correlation, regression, and a full treatment of time series models from moving averages through SARIMA and VAR, closing with the Gaussian mixture model and the expectation-maximisation algorithm.
The exam-relevant habits this course is trying to build: know which distribution a statistic follows and why, state the assumption before you state the test, and read a conclusion as a statement about a population rather than about the sample in front of you.
Each session carries a lab tag showing which of the handout's six labs it prepares you for. The lab sheets themselves sit with your instructor: this course carries the explanation, not the lab submission.
Every module in this course is a different filling of the same five slots. Descriptive statistics, Bayes, a t-test, and an ARIMA forecast all answer the same five questions in the same order.
The official handout cites three textbooks. Source and Source were supplied as PDFs; Source was not — it is Kaptein et al., Statistics for Data Scientists (Springer, 2022) — and the handout cites it for 10 of the 16 contact sessions. So rather than leave you guessing, this table resolves every citation to the Source or Source section that covers the same ground, with the printed page number. Page numbers are the books’ printed numbers, not PDF page indices.
| Module | Handout cites | Read it here |
|---|---|---|
| CS-1 | Source & Source | Source (p. 12) · §1.3 Measures of Location (p. 28) · §1.4 Measures of Variability (p. 35) |
| CS-2 | Source & Source | Source Sample Spaces and Events (p. 51) · §2.2 Axioms, Interpretations, Properties (p. 55) · §2.5 Independence (p. 83) |
| CS-3 | Source & Source | Source Conditional Probability (p. 73) · §2.5 Independence (p. 83) |
| CS-4 | Source & Source | Source Conditional Probability (p. 73). Naive Bayes is not in Source — it is the ML course's S11 material; we cross-link it rather than re-teach it. |
| CS-5 | Source & Source | Source Random Variables (p. 93) · §3.3 Expected Values (p. 106) · §5.1 Jointly Distributed Random Variables (p. 194) · §5.2 Covariance and Correlation (p. 206) |
| CS-6 | Source & Source | Source Binomial (p. 114) · §3.6 Poisson (p. 128) · §4.3 The Normal Distribution (p. 152) · §4.4 Exponential and Gamma (p. 165). t / F / chi-square are introduced in Source (p. 285). |
| CS-7 | Source & Source | Source Populations, Samples, and Processes (p. 2) · §5.4 The Distribution of the Sample Mean (p. 223) · §7.1 Basic Properties of Confidence Intervals (p. 268) |
| CS-8 | — | Mid-semester examination. Revision across CS-1 to CS-7 plus a mock paper. |
| CS-9 | Source & Source | Source Hypotheses and Test Procedures (p. 301) · §8.2 Tests About a Population Mean (p. 310) · §8.3 Tests Concerning a Population Proportion (p. 323) · §8.4 P-Values (p. 328) · §9 Two-Sample Inference (p. 345) |
| CS-10 | Source & Source | Source Methods of Point Estimation (p. 255) · Ch 10 The Analysis of Variance (p. 391) · Ch 11 Multifactor ANOVA (p. 419) |
| CS-11 | Source & Source | Source Expected Values, Covariance, and Correlation (p. 206) · Ch 12 Simple Linear Regression and Correlation (p. 468) |
| CS-12 | Source | Source Some Simple Time Series Models (p. 6) · §1.5 Estimation and Elimination of Trend and Seasonal Components (p. 20) |
| CS-13 | Source | Source Introduction to ARMA Processes (p. 47) · Ch 3 ARMA Models (p. 73) · §3.3 Forecasting ARMA Processes (p. 87) |
| CS-14 | Source | Source ARMA(p, q) Processes (p. 73) · §5.1 Preliminary Estimation (p. 122) · §5.2 Maximum Likelihood Estimation (p. 139) · §5.5 Order Selection (p. 149). SARIMAX / VAR / VARMAX are not in Source — they are flagged as extensions on the module page. |
| CS-15 | Class Notes | Class notes only — outside the scope of both Source and Source. This is also the ML course's S14 (unsupervised) territory; we link rather than duplicate. |
| CS-16 | — | Revision of the syllabus. No single source chapter — the module revisits the exam question style across all six modules. |
Two places in the handout point outside both textbooks: Naive Bayes (CS-4) and GMM / EM (CS-15) are class-notes material. Both are covered properly in the Machine Learning course (S11 and S14), so we cross-link rather than re-teach them thinly here.
Open any module to get the from-scratch ramp, a formal treatment, a cheatsheet, and a practice set. The plate groups below follow the handout’s part structure — CS-8 is the mid-semester pivot where everything from Modules 1–4 has to hold together under exam conditions.
Same modules, ungrouped — for skim-review or to jump straight to a module number.
Central tendency, measures of variability, symmetric vs asymmetric data, outlier detection, the five-number summary, and a first look at probability.
Sample spaces and events, the three axioms, mutually exclusive vs independent events, and problem solving on all three.
Conditional probability given an event, independence as a special case, and the law of total probability.
Bayes' theorem with a full proof, the base-rate problem it solves, and the naive Bayes concept as a classifier.
Discrete and continuous random variables, expectation, mean and variance of a single RV, and joint distributions with covariance.
Bernoulli, binomial, Poisson, and normal distributions, then an introduction to the t, F, and chi-square distributions.
Random and stratified sampling, the sampling distribution of the mean, the central limit theorem, interval estimation, and confidence level.
No new content. A structured revision pass over CS-1 to CS-7, then a mock paper in the format the mid-sem actually uses — 20 questions, 60 minutes, one marking scheme at the end.
Mean-based and proportion-based tests: one mean, two means and one proportion, with small- and large-sample variants wherever each applies. Comparing more than two proportions is Devore’s §9.4 and is not on this page.
Maximum likelihood estimation as a general principle, then ANOVA — single factor and two-factor — for comparing more than two means.
Covariance and correlation, the least-squares line, and how to read a regression model summary without lying to yourself about R².
What makes a time series different, its components, and the basic and weighted moving-average models.
Stationarity, the autocorrelation function, and the AR, ARMA, and ARIMA models — with the ACF/PACF plots that tell you which order to pick.
Seasonal ARIMA, the SARIMAX and VAR extensions, and simple exponential smoothing — plus how to choose an order without fooling yourself.
The Gaussian mixture model, the expectation-maximisation algorithm, and why EM is the general pattern behind k-means and behind latent-variable models everywhere else in your degree.
A full pass over the whole course against the exam question styles — what each module is actually allowed to be asked, and where the marks are.
For course navigation and interactive practice, please use the live app. Offline PDF generation is disabled.