This preview shows pages 1–8. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: σ 2 = Var X i is ﬁnite then the distribution of √ n ( ¯ X nμ ) converges, as n → ∞ , to the normal N (0 ,σ 2 ) distribution. Conclusion : the diﬀerence  ¯ X nμ  ≈ 1 √ n . The CLT is used in diﬀerent forms: • For large n X 1 + X 2 + ... + X nnμ σ √ n has, approximately, the standard normal N (0 , 1) distribution. • P X 1 + X 2 + ... + X nnμ σ √ n ≤ x ! ≈ Φ( x ) if n is large. Rule of thumb: n ≥ 30. • A loose way to formulate the CLT: for large n , X 1 + ... + X n has approximately the normal N ( nμ,nσ 2 ) distribution . • If the variance of X 1 ,X 2 ,... is inﬁnite, then the Central Limit Theorem may still hold, but  ¯ X nμ  ≈ 1 n 11 /α for some 1 < α < 2. • for large n , X 1 + ... + X n has approximately the αstable stable distribution , 1 < α < 2....
View
Full
Document
This note was uploaded on 12/03/2010 for the course OR 3500 at Cornell.
 '09
 SAMORODNITSKY

Click to edit the document details