\((X_{i})\) with CDF \((F_i)\).
Convergence in distribution to X
Aka weak convergence. If \(\forall x: \lim_{n \to \infty} F_n(x) = F(x)\), then \(X_n \to^{d} X\). Comments about limit of CDF’s.
Convergence in probability to X
If \(\forall \eps: \lim_{n \to \infty}Pr(|X_n - X| > \eps) = 0\), say \(X_n \to^p c\); so limit of sequence of probabilities. Probability of deviation from \(X\) grows smaller and smaller, but doesn’t necessarily hit 0. Eg: Weak law of large numbers.
Implies convergence in distribution.
Almost sure convergence
\(Pr(\lim_{n \to \infty}(X_n = X)) = 1\): Note that \(\lim\) is inside Pr(); so limit of sequence of boolean events. Eventually, \(X_n\) will behave exactly like X. Eg: Animal’s daily consumption will some day hit 0 and stay 0.
Implies convergence in probability.
Sure convergence
All \(X_n\) are over exactly same sample space \(\gO\). \(\forall w \in \gO: \lim_{n \to \infty} X_n(w) = X(w)\). Implies all other forms of convergence.