Deeper Learning

[Statistics 110] Lecture 7: Gambler's Ruin and Random Variables 본문

Statistics & Math/Statistics 110: Probability

[Statistics 110] Lecture 7: Gambler's Ruin and Random Variables

Dlaiml 2022. 11. 30. 23:23

Gambler’s Ruin

두 도박사 A, B가 1달러를 걸고 게임을 반복. A가 이길 확률은 $p$, B가 이길 확률은 $1-p$. B가 파산하여 A확률

$p_i = P(A \space wins \space game|A \space starts \space at \space \$i)$로 정의

$p_i = pp_{i+1} + qp_{i=1}$ (1 ≤ i ≤ N-1)

이전 스텝에서 이겨서 현재 위치에 도달했거나 다음 스텝에서 져서 현재 위치에 도달했기 때문에 위와 같이 표현이 가능 [계차방정식(difference equation)]

A가 이길확률이 0.49라면 N이 100일때 0.02 확률로 승리

 

확률변수

확률변수는 Sample space → 실수 공간을 매핑해주는 함수

확률변수 x가 베르누이 분포를 따른다는 것은 x가 2개의 값 0,1만을 가지고 $P(X=1)=p,P(X=0)=1-p$호

확률분포는 곧 확률변수의 청사진, 확률변수가 어떤 값을 가질지를 모두 보여준다

이항분포(n,p)는 n번의 독립적인 베르누이(p) 시행에서 성공 횟수의 분포

 

Reference

[0] https://www.youtube.com/playlist?list=PL2SOU6wwxB0uwwH80KTQ6ht66KWxbzTIo 

 

Statistics 110: Probability

Statistics 110 (Probability) has been taught at Harvard University by Joe Blitzstein (Professor of the Practice in Statistics, Harvard University) each year ...

www.youtube.com

[1] https://www.boostcourse.org/ai152/lecture/30900?isDesc=false 

 

[하버드] 확률론 기초: Statistics 110

부스트코스 무료 강의

www.boostcourse.org

 

Comments