Data Science/ํ†ต๊ณ„

์ด๋ฒคํŠธ A์˜ ์—ฌ์ง‘ํ•ฉ(Complements) ์€ A'์œผ๋กœ ํ‘œ๊ธฐํ•˜๋ฉฐ, A๊ฐ€ ๋ฐœ์ƒํ•˜์ง€ ์•Š์„ ๋ชจ๋“  ํ™•๋ฅ ์„ ์ผ์ปซ๋Š”๋‹ค. (A^c๋กœ ํ‘œ๊ธฐํ•˜๊ธฐ๋„ ํ•ฉ๋‹ˆ๋‹ค.)์ฆ‰, P(A) + P(A') = 1 ์ด๊ณ , ์ด ๋ง์€ P(A') = 1 - P(A) ์ด๋‹ค.๋˜ํ•œ P((A')') = P(A)์ด๋‹ค. (์—ฌ์ง‘ํ•ฉ์˜ ์—ฌ์ง‘ํ•ฉ์€ ์›๋ณธ ์ง‘ํ•ฉ) ์—ฌ์ง‘ํ•ฉ์˜ ์ ์ ˆํ•œ ํ™œ์šฉ์€ ๊ณ„์‚ฐ์„ ๋” ์šฉ์ดํ•˜๊ฒŒ ํ•œ๋‹ค.P(A) = ์ฃผ์‚ฌ์œ„์—์„œ 1, 2, 3, 4, 6์ด ๋‚˜์˜ฌ ํ™•๋ฅ ์ด๋ผ๊ณ  ํ•ด๋ณด์ž.P(A) = 1- P(A')์ด๊ณ , ์ด ๋ง์˜ ๋œป์€ P(A) = 1 - (5๊ฐ€ ๋‚˜์˜ฌ ํ™•๋ฅ ) ์ด๋‹ค.5๊ฐ€ ๋‚˜์˜ฌ ํ™•๋ฅ ์€ 1/6์ด๊ธฐ ๋•Œ๋ฌธ์—, P(A) = 1 - 1/6 = 5/6์ด๋‹ค.  1~6๊นŒ์ง€ 5๋ฅผ ์ œ์™ธํ•˜๊ณ  ๊ฐ ์ˆซ์ž๊ฐ€ ๋‚˜์˜ฌ ํ™•๋ฅ ์„ ๊ตฌํ•ด ๋”ํ•˜๋Š”๊ฒƒ๋ณด๋‹ค 5๊ฐ€ ๋‚˜์˜ฌ ํ™•๋ฅ ์„ ๊ตฌํ•ด์„œ 1์—์„œ ๋นผ๋Š”๊ฒŒ ๋”์šฑ ํŽธ๋ฆฌํ•˜๋‹ค.
2๊ฐœ์˜ ์ฃผ์‚ฌ์œ„๋ฅผ ๋˜์ ธ์„œ ๋‚˜์˜จ ๊ฐ’์˜ ํ•ฉ์˜ ํ™•๋ฅ ์„ ๊ตฌํ•˜๊ณ  ์‹ถ๋‹ค๊ณ  ํ•ด๋ณด๊ฒ ์Šต๋‹ˆ๋‹ค. ๋‚˜์˜ค๋Š” ๊ฐ’๋“ค์„ ํ‘œ๋กœ ๋งŒ๋“ค๋ฉด ๋‹ค์Œ๊ณผ ๊ฐ™์ด ๋˜๊ฒ ์ฃ ?(1, 1) = 2(1, 2) = 3(1, 3) = 4(1, 4) = 5(1, 5) = 6(1, 6) = 7(2, 1) = 3(2, 2) = 4(2, 3) = 5(2, 4) = 6(2, 5) = 7(2, 6) = 8(3, 1) = 4(3, 2) = 5(3, 3) = 6(3, 4) = 7(3, 5) = 8(3, 6) = 9(4, 1) = 5(4, 2) = 6(4, 3) = 7(4, 4) = 8(4, 5) = 9(4, 6) = 10(5, 1) = 6(5, 2) = 7(5, 3) = 8(5, 4) = 9(5, 5) = 10(5, 6) = 11(6, 1) = 7(6, 2) = 8(6, 3) ..
Expected Value๋ž€, ์šฐ๋ฆฌ๊ฐ€ ์‹คํ—˜์„ ์—ฌ๋Ÿฌ๋ฒˆ ๋ฐ˜๋ณตํ–ˆ์„ ๋•Œ, ๊ด€์ธก๋  ๊ฒฐ๊ณผ์˜ ํ‰๊ท ์œผ๋กœ ์˜ˆ์ธกํ•˜๋Š” ๊ฐ’ ์ž…๋‹ˆ๋‹ค.์—ฌ๊ธฐ์„œ ์‹คํ—˜(experiment)์˜ ์ •ํ™•ํ•œ ์ •์˜๋ฅผ ์•Œ๊ณ  ๊ฐ€์•ผ๊ฒ ์ฃ ? ๊ฐ€๋ น ์šฐ๋ฆฌ๊ฐ€ ๋™์ „์„ ๋˜์ ธ์„œ ์•ž๋ฉด์ด ๋‚˜์˜ฌ ํ™•๋ฅ ์ธ P(A)๋ฅผ ๋ชจ๋ฅธ๋‹ค๊ณ  ๊ฐ€์ •ํ•ด๋ด…์‹œ๋‹ค.๊ทธ๋ž˜์„œ ์šฐ๋ฆฌ๋Š” ์ฝ”์ธ ํ† ์Šค๋ฅผ ๋งŽ์ด ๋ฐ˜๋ณตํ•ด์„œ ๋‚˜์˜จ ๊ฒฐ๊ณผ๋ฅผ ๊ด€์ฐฐํ•ด์„œ, ์ด ๊ฐ’์˜ ํ‰๊ท ์„ ๋‚ผ ๊ฒƒ ์ž…๋‹ˆ๋‹ค.ํ•œ๋ฒˆ ๋˜์ ธ์„œ ๊ฒฐ๊ณผ๋ฅผ ๊ด€์ธกํ•˜๋Š”, ์ด ํ•œ๋ฒˆ์˜ ํ–‰์œ„๋ฅผ ์šฐ๋ฆฌ๋Š” ์‹œ๋„(trial)์ด๋ผ๊ณ  ๋ถ€๋ฆ…๋‹ˆ๋‹ค.๊ทธ๋ฆฌ๊ณ  ์ด ์‹œ๋„๋“ค์ด ์—ฌ๋Ÿฌ๋ฒˆ ๋ฐ˜๋ณต๋˜๋Š” ๊ณผ์ •์„ ์‹คํ—˜์œผ๋กœ ์ •์˜ํ•ฉ๋‹ˆ๋‹ค.  ์˜ˆ๋ฅผ ๋“ค์–ด ์šฐ๋ฆฌ๊ฐ€ 20๋ฒˆ ๋™์ „์„ ๋˜์ ธ์„œ ๊ด€์ฐฐํ•œ๋‹ค๊ณ  ํ–ˆ์„ ๋•Œ,์ด๋Š” '20๋ฒˆ์˜ ๊ฐœ๋ณ„์ ์ธ ์‹œ๋„๊ฐ€ ์žˆ๋Š” 1๋ฒˆ์˜ ์‹คํ—˜' ์ž…๋‹ˆ๋‹ค.  Experimental Probabilities | Theoretical Probabil..
ํ™•๋ฅ ์ด๋ž€ ๊ทธ ์ •์˜์—์„œ ์ด๋ฏธ ์•„์‹œ๋‹ค์‹œํ”ผ, ์–ด๋–ค ํŠน์ • ์ด๋ฒคํŠธ x๊ฐ€ ๋ฐœ์ƒํ•  ๊ฐ€๋Šฅ์„ฑ์„ ์ˆซ์ž๋กœ ๋‚˜ํƒ€๋‚ธ ๊ฒƒ ์ž…๋‹ˆ๋‹ค.์šฐ๋ฆฌ๋Š” ๋ฏธ๋ž˜์— ๋ฒŒ์–ด์งˆ ์ผ์˜ ํ™•๋ฅ ๋“ค์„ ์ˆ˜์ ์œผ๋กœ ๋น„๊ตํ•ด์„œ, ๋”์šฑ ๋†’์€ ํ™•๋ฅ ์ด ์–ด๋–ค ๊ฒƒ ์ธ์ง€๋ฅผ ์•Œ์•„๋ƒ„์œผ๋กœ์จ ๋ฏธ๋ž˜๋ฅผ ์˜ˆ์ธกํ•˜๊ณ ์ž ํ•ฉ๋‹ˆ๋‹ค. ์šฐ๋ฆฌ๋Š” ์–ด๋– ํ•œ ์‚ฌ๊ฑด x๊ฐ€ ์ ˆ๋Œ€๋กœ ๋ฐœ์ƒํ•˜์ง€ ์•Š๋Š”๋‹ค๋Š” ๊ฒƒ์„ ํ™•๋ฅ ์ด 0์ด๋‹ค. ๋ผ๊ณ  ํ‘œํ˜„ํ•ฉ๋‹ˆ๋‹ค.๊ทธ๋ฆฌ๊ณ  ์šฐ๋ฆฌ๋Š” ์–ด๋– ํ•œ ์‚ฌ๊ฑด x๊ฐ€ ๋ฌด์กฐ๊ฑด ๋ฐœ์ƒํ•œ๋‹ค๋Š” ๊ฒƒ์„ ํ™•๋ฅ ์„ 1์ด๋‹ค. ๋ผ๊ณ  ํ‘œํ˜„ํ•ฉ๋‹ˆ๋‹ค.์ฆ‰ ๋ชจ๋“  ํ™•๋ฅ ์€ 0~1์˜ ๊ฐ’์„ ๊ฐ€์ง€๊ณ , ๋ณดํ†ต ์šฐ๋ฆฌ์˜ ๋ฐ์ดํ„ฐ๋Š” 0๊ณผ 1์€ ์•„๋‹ˆ๊ฒ ์ฃ ? ๊ณ„์‚ฐ๊ณผ ๋น„๊ต์˜ ํŽธ์˜์„ฑ์„ ์œ„ํ•ด, ์šฐ๋ฆฌ๋Š” 30%, 1/5, ์ด๋Ÿฐ ์ˆ˜๋กœ ํ™•๋ฅ ์„ ํ‘œํ˜„ํ•˜๊ธฐ๋ณด๋‹ค 0.2, 0.53๊ณผ ๊ฐ™์€ ์†Œ์ˆ˜๋กœ ํ‘œํ˜„ํ•˜๊ธฐ๋ฅผ ์„ ํ˜ธํ•ฉ๋‹ˆ๋‹ค.  ํŠน์ • ์ด๋ฒคํŠธ x์— ๋Œ€ํ•ด์„œ ์šฐ๋ฆฌ๋Š”,x๊ฐ€ ์ผ์–ด๋‚  ํ™•๋ฅ ์„ P(x)๋กœ ํ‘œ๊ธฐํ•˜๊ณ ,P(x)..
Chan Lee
'Data Science/ํ†ต๊ณ„' ์นดํ…Œ๊ณ ๋ฆฌ์˜ ๊ธ€ ๋ชฉ๋ก (2 Page)