Plain-language explanation.
Probability theory is the mathematical framework for quantifying uncertainty and randomness. It provides the rigorous foundations for statistics, machine learning, finance, physics, and every field that deals with uncertain or variable outcomes.
Core concepts and standard treatment.
Core probability covers sample spaces and events (sigma-algebras — measure-theoretic foundations), axioms of probability (Kolmogorov axioms), conditional probability and independence (Bayes' theorem, chain rule), random variables (discrete and continuous — PMF, PDF, CDF), expectation and variance (linearity of expectation, Var(X+Y) for independent variables), key distributions (binomial, Poisson, geometric, normal, exponential, uniform, beta, gamma), and transformations of random variables.
Deeper theory, debates and edge cases.
Advanced probability theory covers the law of large numbers (weak vs strong LLN — convergence in probability vs almost sure convergence), central limit theorem (Lindeberg conditions, Berry-Esseen bounds), generating functions (MGF, characteristic function — uniqueness theorem), stochastic processes (Markov chains — transition matrices, stationary distributions, mixing time; Brownian motion — the Wiener process; Poisson processes), martingales (optional stopping theorem — applications in gambling and finance), and large deviation theory (Cramér's theorem).
How it is applied in practice.
At the quant researcher, ML engineer, and risk manager level, probability theory underpins: option pricing (risk-neutral measure, Girsanov's theorem — Black-Scholes via risk-neutral pricing); reinforcement learning (Markov decision processes, Bellman equations, policy gradient — REINFORCE algorithm); Bayesian deep learning (variational inference, ELBO — VAEs); statistical physics (Ising model, partition functions); and reliability engineering (failure rate, hazard function, reliability block diagrams — aerospace and nuclear).