1 · Foundations
What statistics is, why ML needs it, and the population-vs-sample distinction every experiment rests on.
●Track 3 · live
Statistics & Probability
The numerical reasoning every model rests on.
Descriptive statistics, distributions, hypothesis testing, ANOVA, Bayes — with scipy and statsmodels running in your browser. Each topic has a 'why this matters in ML' connection and a link to the matching MML chapter for the math.
29 / 30 lessons8 chapters
2 · Descriptive statistics
Mean, median, variance, percentiles, correlation — the toolkit for summarising any dataset in five numbers and a picture.
3 · Probability foundations
Sum and product rules, PMF/PDF/CDF, and the six distributions that cover almost every ML task.
4 · Discrete distributions
Bernoulli, Binomial, Poisson — the count-and-yes/no trio that underlies binary classification, A/B testing, and count regression.
5 · Continuous distributions
Normal, Z-score, Uniform, Log-normal, Power-law — the continuous distributions every ML model assumes under the hood.
6 · Inference foundations
Central Limit Theorem, estimation with MLE and confidence intervals, hypothesis testing, and what the p-value really means.
7 · Hypothesis tests
Z, t, paired, Type I/II errors, and Bayes — the formal statistical tests every ML experiment relies on.
8 · Multi-group testing
Chi-square for categorical data and ANOVA for comparing 3+ group means — the tools for any multi-variant experiment.