์ธ๊ณต์ง€๋Šฅ์ˆ˜ํ•™ 2

[AI] ๋ชจ๋น„์œจ

๋ชจ๋น„์œจ์˜ ์ถ”์ • ๋ชจ๋น„์œจ: ๋ชจ์ง‘๋‹จ์—์„œ ์–ด๋–ค ํŠน์„ฑ์„ ๊ฐ–๋Š” ์ง‘๋‹จ์˜ ๋น„์œจ ์ข…๋ฅ˜ 1. ์ ์ถ”์ • ํ™•๋ฅ ๋ณ€์ˆ˜ X: n๊ฐœ์˜ ํ‘œ๋ณธ์—์„œ ํŠน์ • ์†์„ฑ์„ ๊ฐ–๋Š” ํ‘œ๋ณธ์˜ ๊ฐœ์ˆ˜ ๋ชจ๋น„์œจ ๐‘์˜ ์ ์ถ”์ •๋Ÿ‰ $\hat๐‘ = \frac{X}{n}$ ์˜ˆ์ œ ๋Œ€ํ•™๊ต 1 ํ•™๋…„์ƒ์˜ ํก์—ฐ์œจ์„ ์กฐ์‚ฌํ•˜๊ธฐ ์œ„ํ•ด 150 ๋ช…์„ ๋žœ๋คํ•˜๊ฒŒ ์„ ํƒํ•˜์—ฌ ํก์—ฐ์—ฌ๋ถ€๋ฅผ ์กฐ์‚ฌํ•˜์˜€๋‹ค . ์ด ์ค‘ 48 ๋ช…์ด ํก์—ฐ์„ ํ•˜๊ณ  ์žˆ์—ˆ๋‹ค . ์ด ๋Œ€ํ•™๊ต 1 ํ•™๋…„์ƒ์˜ ํก์—ฐ์œจ์˜ ํ‰๊ท ์„ ์ ์ถ”์ •ํ•˜์‹œ์˜ค . ๐‘› = 150 , ๐‘‹ = 48 $\hat๐‘ = \frac{๐‘‹}{๐‘›} = \frac{48}{150} = 0.32$ ํ‰๊ท ํก์—ฐ์œจ์„ 32% ๋กœ ์ถ”์ •๋จ 2. ๊ตฌ๊ฐ„์ถ”์ • ๐‘›์ด ์ถฉ๋ถ„ํžˆ ํด ๋•Œ, $n \hat p > 5, n(1 - \hat p) > 5$ ์ผ ๋•Œ๋ฅผ ์˜๋ฏธ X~N(np, np(1 - p)) ํ™•๋ฅ ๋ณ€์ˆ˜ X์˜ ํ‘œ์ค€ํ™”..

CS/AI 2023.06.12

[AI] ๋ชจํ‰๊ท 

๋ชจํ‰๊ท  μ ๋ชจํ‰๊ท ์˜ ์ถ”์ •์— ๋Œ€ํ•ด ์ •๋ฆฌํ•ด ๋ณด๋ ค๊ณ  ํ•ฉ๋‹ˆ๋‹ค. ์ˆ˜์—…์ด ๊ฐ€๋ฉด ๊ฐˆ์ˆ˜๋ก ์–ด๋ ค์›Œ์ง€๋„ค์š”. ์š”์ฆ˜๋“ค์–ด ํ•™์ฐฝ์‹œ์ ˆ ์ˆ˜ํ•™ ๊ณต๋ถ€๋ฅผ ํ•˜์ง€ ์•Š์•˜๋˜ ์ € ์ž์‹ ์ด ํ›„ํšŒ๋˜๋Š” ์‹œ๊ฐ„์„ ์ž์ฃผ ๊ฐ–๋Š” ๋“ฏ ํ•ฉ๋‹ˆ๋‹ค. ๋’ค์ฒ˜์ง„ ๋งŒํผ ๊พธ์ค€ํžˆ ํ•ด๋‚˜๊ฐ€์•ผํ•  ๋“ฏ ํ•ฉ๋‹ˆ๋‹ค. ๊ตญ๋ฏผ๋Œ€ ์ธ๊ณต์ง€๋Šฅ์ˆ˜ํ•™ 13์ฃผ์ฐจ ์ˆ˜์—… ์ž๋ฃŒ๋ฅผ ๋ฐ”ํƒ•์œผ๋กœ ์ž‘์„ฑ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ๋ชจํ‰๊ท ์˜ ์ถ”์ • ๋ชจํ‰๊ท ์ด ์ •๊ทœ๋ถ„ํฌ์ธ ๊ฒฝ์šฐ ํ‘œ๋ณธํ‰๊ท ์„ ์‚ฌ์šฉ $$ \overline{X} = \frac{X_1+X_2+\dots+X_n}{n}= \frac{\sum_{i=1}^{10}X_1}{n} $$ ์ข…๋ฅ˜ 1. ์ ์ถ”์ • ํ‘œ๋ณธํ‰๊ท ์ด ์  ์ถ”์ „๊ฐ’ (์ถ”์ •๋Ÿ‰)์ด ๋จ import numpy as np ## ํ‘œ๋ณธ์ถ”์ถœ samples = [9,4,0,8,1,3,7,8,4,2] ## ํ‘œ๋ณธ์˜ ํ‰๊ท  = ๋ชจ์ง‘๋‹จ์˜ ์ ์ถ”์ • ๊ฐ’ print(np.mea..

CS/AI 2023.06.07
728x90
๋ฐ˜์‘ํ˜•