UA MATH563 The Central Limit Theorem of the Mathematical Foundation of Probability Theory 22 The Portmanteau Theorem of Weak Convergence in Metric Probability Space
Now we discuss weak convergence in the metric space, assuming (Ω, d) (\Omega,d)( Oh ,d ) is a metric space,(Ω, F, P) (\Omega,\mathcal{F},P)( Oh ,F,P ) is a probability space,X n, X X_n, XXn,X is defined inΩ \OmegaRandom variables on Ω , their distribution isμ n, μ \mu_n,\muμn,μ。
Portmanteau's theorem
With regard to convergence in distribution, the following statements are equivalent:
- X n → d X X_n \to_d X Xn→dX
- For any open set GGG, lim inf P ( X n ∈ G ) ≥ P ( X ∈ G ) \liminf P(X_n \in G) \ge P(X \in G) l i minfP(Xn∈G)≥P(X∈G)
- For any closed set KKK, lim sup P ( X n ∈ K ) ≤ P ( X ∈ K ) \limsup P(X_n \in K) \le P(X \in K) l i msupP(Xn∈K)≤P(X∈K)
- AA for any setA , ifP (X ∈ ∂ A) = 0 P(X \in \partial A) = 0P(X∈∂A)=0,则 lim P ( X n ∈ A ) = P ( X ∈ A ) \lim P(X_n \in A) = P(X \in A) limP(Xn∈A)=P(X∈A)
Regarding weak convergence, the following statements are equivalent:
- μ n ⇒ μ \ mu_n \ Rightarrow \ mu μn⇒μ
- For any open set GGG,lim inf μ n (G) ≥ μ (G) \ liminf \ mu_n (G) \ ge \ mu (G)l i minfμn(G)≥μ ( G )
- For any closed set KKK,lim soup μ n (K) ≤ μ (K) \ limsup \ mu_n (K) \ le \ mu (K)l i msupμn(K)≤μ ( K )
- AA for any setA , ifμ (∂ A) = 0 \mu(\partial A) = 0μ ( ∂ A )=0,则 μ ( A n ) → μ ( A ) \mu(A_n) \to \mu(A) μ ( An)→μ ( A )
The path of proof is 1 ⇒ 3 ⇒ 2 ⇒ 4 ⇒ 1 1 \Rightarrow 3 \Rightarrow 2 \Rightarrow 4 \Rightarrow 11⇒3⇒2⇒4⇒1. Post a Durrett certificate