๋ณธ๋ฌธ ๋ฐ”๋กœ๊ฐ€๊ธฐ

๐Ÿ’ซ ์ˆ˜ํ•™

ํ†ต๊ณ„ํ•™ 3์ฃผ์ฐจ - ํšŒ๊ท€๋ถ„์„ ํ†ต๊ณ„์  ์ถ”์ •์„ ํ•˜๋Š” ์ด์œ ๋Š” ๋ฌด์—‡์ธ๊ฐ€? ์–ด๋–ค ๋ชจ์ง‘๋‹จ์„ ๋Œ€์ƒ์œผ๋กœ ์‹คํ—˜์ด๋‚˜ ์กฐ์‚ฌ๋ฅผ ํ• ๋•Œ, ์‹œ๊ฐ„๊ณผ ๋น„์šฉ์˜ ์ œ์•ฝ์ด ์žˆ์–ด ์ „์ˆ˜์กฐ์‚ฌ๋ฅผ ํ•˜๊ธฐ ํž˜๋“œ๋ฏ€๋กœ ํ‘œ๋ณธ์„ ๋ฝ‘์•„ ์กฐ์‚ฌ ๋ฐ ์‹คํ—˜์„ ์ง„ํ–‰ํ•˜์—ฌ ๋ชจ์ง‘๋‹จ์˜ ํ™•๋ฅ ๋ถ„ํฌ๋ฅผ ์ถ”์ •ํ•˜๋Š” ํ†ต๊ณ„์  ์ถ”์ •์„ ํ•œ๋‹ค. ์ ์ถ”์ •๊ณผ ๊ตฌ๊ฐ„์ถ”์ •์˜ ์ฐจ์ด๋ฅผ ์„ค๋ช…ํ•˜์‹œ์˜ค. (๊ตฌ๊ฐ„์ถ”์ •์˜ ๊ฒฝ์šฐ, ์‹ ๋ขฐ๊ตฌ๊ฐ„์˜ ์ •์˜์™€ ํ•จ๊ป˜ ์„ค๋ช…ํ•˜์‹œ์˜ค) ์ ์ถ”์ •์ด๋ž€ ์ˆ˜์น˜์  ์ถ”์ •์น˜๋ฅผ ์˜ˆ์ธกํ•˜๋Š” ๊ฒƒ์ด๋‹ค. ๊ตฌ๊ฐ„์ถ”์ •์ด๋ž€ ์ผ์ • ๊ตฌ๊ฐ„์•ˆ์˜ ์ตœ์†Ÿ๊ฐ’๊ณผ ์ตœ๋Œ“๊ฐ’ ์‚ฌ์ด์˜ ๊ฐ’์ด๋ผ๊ณ  ์ถ”์ •ํ•˜๋Š” ๊ฒƒ์ด๋‹ค. ๋Œ€ํ‘œ์ ์ธ ๊ตฌ๊ฐ„์ถ”์ • ๋ฐฉ๋ฒ•์€ ์‹ ๋ขฐ๊ตฌ๊ฐ„์žˆ๋‹ค. ์‹ ๋ขฐ๊ตฌ๊ฐ„์€ ํ‘œ๋ณธ ํ†ต๊ณ„๋Ÿ‰์—์„œ ํŒŒ์ƒ๋˜์–ด ์•Œ ์ˆ˜ ์—†๋Š” ๋ชจ์ง‘๋‹จ ๋ชจ์ˆ˜ ๊ฐ’์ด ํฌํ•จ๋  ๊ฐ€๋Šฅ์„ฑ์ด ์žˆ๋Š” ๊ฐ’์˜ ๋ฒ”์œ„์ด๋‹ค. ๋…๋ฆฝ๋ณ€์ˆ˜์™€ ์ข…์†๋ณ€์ˆ˜ ๊ฐ๊ฐ์— ๋Œ€ํ•ด ์˜ˆ์‹œ๋ฅผ ๋“ค์–ด ์„ค๋ช…ํ•˜์‹œ์˜ค. ๋…๋ฆฝ๋ณ€์ˆ˜๋ž€ ์ข…์†๋ณ€์ˆ˜๋ฅผ ์„ค๋ช…ํ•ด์ฃผ๋Š” ๋ณ€์ˆ˜๋ฅผ ๋งํ•œ๋‹ค. ์ข…์†๋ณ€์ˆ˜๋ž€ ๋…๋ฆฝ๋ณ€์ˆ˜์— ์˜.. ๋”๋ณด๊ธฐ
๋ฒกํ„ฐ vectors What is the Vector? Vector Scalars ( magnitude | size ) AND direction magnitude | size Equivalent vectors ์„œ๋กœ size์™€ direction์ด ๊ฐ™์€ ๋ฒกํ„ฐ๋ฅผ ์„œ๋กœ ๊ฐ™์€ ๋ฒกํ„ฐ๋ผ ํ•œ๋‹ค. Components of Vectors x์˜ ๋ณ€ํ™”๋Ÿ‰, y์˜ ๋ณ€ํ™”๋Ÿ‰์„ ๊ฐ๊ฐ vector์˜ component๋ผ ํ•œ๋‹ค. ๋”๋ณด๊ธฐ
ํ†ต๊ณ„ํ•™์ž…๋ฌธ 2์ฃผ์ฐจ - ๋ฐฐ๋ฐ˜์‚ฌ๊ฑด ์—ฌ์‚ฌ๊ฑด ๋…๋ฆฝ์‚ฌ๊ฑด ๋žœ๋ค ํ”„๋กœ์„ธ์Šค(Random Process)๋ž€ ๋ฌด์—‡์ธ๊ฐ€? ์–ด๋–ค ์‹คํ—˜๊ฒฐ๊ณผ๋ฅผ ํ•˜๋‚˜์˜ ํ•จ์ˆ˜์— ๋Œ€์‘์‹œํ‚ค๋Š” ๊ฒƒ ์•ž๋ฉด๊ณผ ๋’ท๋ฉด์ด ๋‚˜์˜ฌ ํ™•๋ฅ ์ด ๊ฐ๊ฐ 1/2์ธ ๋™์ „์„ 5๋ฒˆ ๋˜์กŒ๋Š”๋ฐ ์•ž๋ฉด์ด 4๋ฒˆ, ๋’ท๋ฉด์ด 1๋ฒˆ ๋‚˜์™”๋‹ค. ๋™์ „์„ ํ•œ๋ฒˆ ๋” ๋˜์กŒ์„ ๋•Œ, ์•ž๋ฉด์ด ๋‚˜์˜ฌ ํ™•๋ฅ ์€? 1/2 ๋ฐฐ๋ฐ˜์‚ฌ๊ฑด(Disjoint events)๊ณผ ์—ฌ์‚ฌ๊ฑด(complementary events)์˜ ์ฐจ์ด๋ฅผ ์„ค๋ช…ํ•˜์‹œ์˜ค. ๋ฐฐ๋ฐ˜์‚ฌ๊ฑด์€ ๋™์‹œ์— ์ผ์–ด๋‚˜์ง€ ์•Š์€ ์‚ฌ๊ฑด. ์‚ฌ๊ฑด A์— ๋Œ€ํ•œ ์ง‘ํ•ฉ์ด ์žˆ์„๋•Œ, ์‚ฌ๊ฑด A ์ง‘ํ•ฉ์— ํฌํ•จ๋˜์–ด์žˆ๋Š” ์‚ฌ๊ฑด๋˜๋Š” ํฌํ•จ๋˜์–ด์žˆ์ง€ ์•Š์€ ์‚ฌ๊ฑด ๋ชจ๋‘ ๋™์‹œ์— ์ผ์–ด๋‚˜์ง€ ์•Š๋Š”๋‹ค๋ฉด (๋…๋ฆฝ์ ์ด๋ผ๋ฉด) ๋ชจ๋‘ ๋ฐฐ๋ฐ˜์‚ฌ๊ฑด์ด๋‹ค. ์—ฌ์‚ฌ๊ฑด์€ ์‚ฌ๊ฑด A์ง‘ํ•ฉ์— ํฌํ•จ๋˜์ง€ ์•Š๋Š” ์‚ฌ๊ฑด. ์ด๋•Œ ์ „์ฒด์ง‘ํ•ฉ์˜ ํ•ฉ์€ 1์ด ๋œ๋‹ค. ๋…๋ฆฝ ์‚ฌ๊ฑด(Independent Event)์ด๋ž€ ๋ฌด์—‡์ธ๊ฐ€? ๋‘ ์‚ฌ๊ฑด์ด .. ๋”๋ณด๊ธฐ
ํ†ต๊ณ„ํ•™ ์ž…๋ฌธ 1์ฃผ์ฐจ - ์ธ๊ณผ๊ด€๊ณ„, ์ƒ๊ด€๊ด€๊ณ„, ์„ค๋ฌธ์กฐ์‚ฌ๋Š” ์™œ ๋ถˆ์™„์ „ํ•  ์ˆ˜ ์žˆ๋Š”๊ฐ€? ์งˆ๋ฌธ1. ํ†ต๊ณ„ํ•™์ด๋ž€ ๋ฌด์—‡์ด๊ณ  ์™œ ์ค‘์š”ํ•˜๋‹ค๊ณ  ์ƒ๊ฐํ•˜๋Š”๊ฐ€? * ํ†ต๊ณ„ํ•™์ด๋ž€ ํ˜„์‹ค์— ์กด์žฌํ•˜๋Š” ์ •๋ณด๋ฅผ ์ •๋Ÿ‰์ ์œผ๋กœ ๋‚˜ํƒ€๋‚ด๊ธฐ ์œ„ํ•œ ๋„๊ตฌ์ด๋‹ค. ํ†ต๊ณ„๋Š” ์ •๋ณด๋ฅผ ๊ฐ๊ด€์„ฑ์„ ๋ณด์žฅํ•ด์ฃผ๊ธฐ ๋•Œ๋ฌธ์— ํ†ต๊ณ„๊ธฐ๋ฐ˜ ์‹œ๊ฐ์ž๋ฃŒ๋ฅผ ์ด์šฉํ•˜๋ฉด ์ƒ๋Œ€๋ฅผ ์„ค๋“ํ•˜๋Š”๋ฐ ๋„์›€์ด ๋œ๋‹ค. ์งˆ๋ฌธ2. ๋ชจ์ง‘๋‹จ (Population)๊ณผ ์ƒ˜ํ”Œ (Sample)์˜ ์ฐจ์ด๋Š”? ์šฐ๋ฆฌ๋Š” ์™œ ์ƒ˜ํ”Œ๋ง์„ ํ•˜๋Š”๊ฐ€? * ๋ชจ์ง‘๋‹จ์ด ์ „์ฒด์ง‘ํ•ฉ์ด๋ผ๋ฉด ์ƒ˜ํ”Œ์€ ๋ถ€๋ถ„์ง‘ํ•ฉ๋‹ˆ๋‹ค. ์–ด๋–ค study์—์„œ ์ฆ๋ช…?๋Œ€๋ณ€? ํ•˜๊ณ ์žํ•˜๋Š” ๋Œ€์ƒ์€ ๋ชจ์ง‘๋‹จ์ด๊ณ , ๊ฑฐ๊ธฐ์„œ ๋Œ€ํ‘œ์„ฑ representative์žˆ๊ฒŒ ๋ฝ‘์•„๋‚ธ ์ž‘์€ ๋ถ€๋ถ„์ง‘ํ•ฉ์„ ์ƒ˜ํ”Œ์ด๋ผ๊ณ  ๋ถ€๋ฅธ๋‹ค. ์งˆ๋ฌธ3. ์ •๋Ÿ‰ ๋ณ€์ˆ˜ (quantitative)์™€ ์ •์„ฑ ๋ณ€์ˆ˜ (qualitative)์˜ ์ฐจ์ด์™€ ๊ฐ๊ฐ์˜ ํŠน์ง•์€ ๋ฌด์—‡์ธ๊ฐ€? * ์ •๋Ÿ‰๋ณ€์ˆ˜๋Š” numerical ๋ณ€์ˆ˜๋ผ๊ณ ๋„ ๋ถ€๋ฅด๋Š”๋ฐ, ์‚ฐ์ˆ ์—ฐ์‚ฐ์ด.. ๋”๋ณด๊ธฐ
Linear Equation and Linear System Scalar: a single number s (- R e.g, 3.8 Vector: ๋ฒกํ„ฐ๋Š” ํฌ๊ธฐ์™€ ๋ฐฉํ–ฅ์„ ๋™์‹œ์— ๋‚˜ํƒ€๋‚ธ๋‹ค. vector indicate magnitude and direction ์†๋„ velocity = 5mpu (ํž˜ magnitude) + East (๋ฐฉํ–ฅ direction) an ordered list of numbers. (an unordered list of numbers: set) - column vector์™€ row vector๊ฐ€ ์žˆ์Œ A vector of n-dimension is usually a column vector n by 1. Thus, a row vector is usually written as its transpose. Matrix: a two-dimensio.. ๋”๋ณด๊ธฐ
ํšŒ๊ท€๋ถ„์„์„ ์œ„ํ•œ ์„ ํ˜•๋ชจ๋ธ, Linear Models for Regression ํšŒ๊ท€๋ถ„์„ - ์ง€๋„ ํ•™์Šต ๋ชฉํ‘œ: ์‹ค์ˆ˜ ๋ฒ”์œ„์˜ - ํ•ด์„๋ ฅ์ด ์ข‹๋‹ค - ํ›ˆ๋ จ ๋ฐ์ดํ„ฐ ๊ด€์ธก๋˜๊ธฐ ์ „์— ๊ณ ์ • ์ฐจ์›์˜ ์ €์ฃผ ๋ฌธ์ œ ์žˆ์Œ bias varience trade off ์˜ˆ์ธก trad off ์ถ”๋ก  - ISIR ์˜ˆ์ธก ์ •ํ™•๋„๊ฐ€ ๋†’์œผ๋ฉด ( ์œ ์—ฐ์„ฑflexivity์ด ๋†’์•„์„œ) ๋ชจ๋ธ์„ ํ•ด์„ํ•  ์ˆ˜ ์žˆ๋Š” ํ•ด์„๋ ฅ์ด ๋‚ฎ์•„์ง€๋Š” ๊ด€๊ณ„ ๊ธฐ์ €ํ•จ์ˆ˜ basis function ฯ•(⋅) - ๊ณต๊ฐ„์„ ๋ฐ”๊ฟ”์ค€๋‹ค? - ํ™œ์„ฑํ™” ํ•จ์ˆ˜๊ฐ™์€ ๋ถ€์—ฐ์„ค๋ช…์ด ๋‹ฌ๋ ค์žˆ๋Š”๊ฑฐ๊ฐ™์€๋ฐ(์•„๋‹˜); ๊ธฐ์ €ํ•จ์ˆ˜์˜ ๋„์ž…์œผ๋กœ ๊ธฐ์กด์—๋Š” x์— ๋Œ€ํ•œ ์„ ํ˜• ์‹์ด์—ˆ๋˜ y(x, w) ํ•จ์ˆ˜๊ฐ€ x์— ๋Œ€ํ•œ ๋น„์„ ํ˜• ํ•จ์ˆ˜๊ฐ€ ๋  ์ˆ˜๋„ ์žˆ๋‹ค. - ๊ธฐ์ €ํ•จ์ˆ˜์— PCA๋ฅผ ๋„ฃ์„ ์ˆ˜๋„ ์žˆ๊ณ  - ์„ ํ˜•ํšŒ๊ท€์˜ ์„ ํ˜•์ด๋ผ๋Š”๊ฒŒ W์— ๋Œ€ํ•ด ์„ ํ˜•์ด๋ผ๋Š” ๊ฒƒ์ด๊ณ , x์— ๋Œ€ํ•ด์„œ๋Š” ์„ ํ˜•์ด ์•„๋‹ˆ์—ฌ๋„ ๋œ๋‹ค. - ๊ธฐ์ €ํ•จ์ˆ˜ ์ „์ฒ˜๋ฆฌํ•ด์ฃผ๋Š”๊ฑฐ๊ฐ™๋„ค n.. ๋”๋ณด๊ธฐ
Designing Studies Introduction to Probability and Data ์ˆ˜์—…์„ ๋“ค์œผ๋ฉฐ ๋‚จ๊ธด ๋…ธํŠธ์ž…๋‹ˆ๋‹ค. Week1: Designing studies 1. Data Basics 2. Observational Studies & Experiments 3. Sampling and sources of bias 4. Experimental Design 5. Random Sample Assignment 1. Data Basics * Observations / Variables / Data matrices * types of variables all variable Numerical (quantitative) Categorical (qualitative) ๋”ํ•˜๊ณ , ๋นผ๊ณ , ํ‰๊ท ์„ ๊ตฌํ•  ์ˆ˜ ์žˆ๋Š” ์ˆ˜์น˜์  ๊ฐ’์„ ๊ฐ€์ง„๋‹ค. take .. ๋”๋ณด๊ธฐ
๊ฒฝ์šฐ์˜ ์ˆ˜ ๊ฒฝ์šฐ์˜ ์ˆ˜๋Š” ๋”ํ•˜๊ณ  ๊ณฑํ•˜๊ณ  ๋นผ๊ณ  ๋‚˜๋ˆ„๋Š” ์‚ฌ์ธก์—ฐ์‚ฐ์„ ํ†ตํ•ด ๊ตฌํ•  ์ˆ˜ ์žˆ๋‹ค.+ (๋”ํ•˜๊ธฐ) A or B; ๋‘ ์‚ฌ๊ฑด A, B๊ฐ€ ๋™์‹œ์— ์ผ์–ด๋‚˜์ง€ ์•Š์„ ๋•Œ, A์˜ ๊ฒฝ์šฐ์˜ ์ˆ˜ +B์˜ ๊ฒฝ์šฐ์˜ ์ˆ˜ ์œ—์˜ท 3๊ฐœ ์ค‘ 1๊ฐœ๋ฅผ ๊ณจ๋ผ ์ž…๋Š” ๊ฒฝ์šฐ์˜ ์ˆ˜1 + 1 + 1 = 3 ๊ฐ€์ง€ * (๊ณฑํ•˜๊ธฐ) A and B; ๋‘ ์‚ฌ๊ฑด A, B๊ฐ€ ๋™์‹œ์— ์ผ์–ด๋‚  ๋•Œ, A์˜ ๊ฒฝ์šฐ์˜ ์ˆ˜ * B์˜ ๊ฒฝ์šฐ์˜ ์ˆ˜ - ๋™์‹œ์— ์œ—์˜ท 3๊ฐœ์™€ ๋ฐ”์ง€ 2๊ฐœ๋ฅผ ๋งค์น˜ํ•˜๋Š” ๊ฒฝ์šฐ์˜ ์ˆ˜ ( 1 + 1 + 1 ) * ( 1 + 1 ) = 3 * 2 = 6 ๊ฐ€์ง€ - ์—ฐ์†ํ•ด์„œa ์—์„œ b๋ฅผ ๊ฑฐ์ณ c๋กœ ๊ฐ€๋Š” ๋ฐฉ๋ฒ•.a ์—์„œ b๋กœ ๊ฐ€๋Š” ๋ฐฉ๋ฒ• 2 ๊ฐ€์ง€ * b์—์„œ c๋กœ ๊ฐ€๋Š” ๋ฐฉ๋ฒ• 3 ๊ฐ€์ง€ ( 1 + 1 ) * ( 1 + 1 + 1 ) = 2 * 3 = 6 - (๋นผ๊ธฐ)์ค‘๋ณต๋˜๋Š” ๊ฒฝ์šฐ์˜ ์ˆ˜๋ฅผ ๋บด์ค€๋‹ค.์ง‘ํ•ฉ.. ๋”๋ณด๊ธฐ