2023. 5. 18. 20:08ㆍ딥러닝 이론/수학
Laplace가 probability의 정의를 이렇게 하였다:
"The probability that A occurs" => Pr(A) = |A|/|S| = |A|/n where S is the sample space, A is a subset of S, and |S| = n. |
요즘은 초/중학교에서도 배우는 내용이겠지만 처음 정의됐을 때만큼은 혁신적이었을 것이다. Probability theory는 원래 도박류 게임을 연구하기 위해 1700~1800년대에는 많은 관심을 끌었다. 지금은 공학에서 필수적으로 배우는 학문 중 하나이다.
정의는 쉽게 이해가 되지만, Counting문제와 비슷하게 많이 까다로울 때가 많다.
문제를 몇개 살펴봐요:
From a standard deck of 52 cards, what is the probability of drawing (a hand of)
(a) three aces and two jacks?
(b) three aces and a pair?
(c) three of one kind and a pair (a full house)?
(a) Since we are getting a hand of cards, it is a set of cards (order does not matter, and without replacement).
We first choose 5 cards from 52 = 52C5. |S| = 52C5
There are 3 aces in total, so we have 4C3. |A| = 4C3
There are 4 jacks in total, so we have 4C2. |B| = 4C2
Pr(|A| intersection |B|) = |A| * |B| / |S| = (4C3)(4C2) / (52C5)
(b) A pair is any 2 cards of the same kind (value).
There are a total of 13 kinds of cards, but we already have 3 aces (one kind of 13), so |C| = (12C1) * (4C2)
Pr(|A| intersection |C|) = |A||C|/|S| = (4C3)(12C1)(4C2)/(52C5)
(c) Three of one kind is basically the 3x option of a pair.
There are again 13 kinds, so three of a kind and a pair (2 of a kind) requires 2 distinct kinds.
(13C1)(4C3) * (12C1)(4C2) / (52C5)
Fin~