profile-img
The merit of an action lies in finishing it to the end.
slide-image
์ „์ฒด ๊ธ€ + 29
list_img
Microaveraging vs. Macroaveraging
user-img iamnotwhale
2023.11.14
์ด์ง„ ๋ถ„๋ฅ˜๊ธฐ์˜ ์„ฑ๋Šฅ์„ ๊ณ„์‚ฐํ•  ๋•Œ ์šฐ๋ฆฌ๋Š” confusion matrix (ํ˜ผ๋™ํ–‰๋ ฌ) ์„ ๊ทธ๋ ค True positive, True negative, False positive, False negative๋ฅผ ํ™•์ธํ•˜๊ณ  ์ด๋ฅผ ํ†ตํ•ด precision, recall, accuracy๋“ฑ์„ ํ™•์ธํ•  ์ˆ˜ ์žˆ๋‹ค. ๊ทธ๋Ÿฐ๋ฐ ๋งŒ์•ฝ ๋ถ„๋ฅ˜ํ•ด์•ผ ํ•  ํด๋ž˜์Šค๊ฐ€ 3๊ฐœ ์ด์ƒ์ด๋ผ๋ฉด ์–ด๋–ป๊ฒŒ ํ•ด์•ผํ• ๊นŒ? ๋‘ ๊ฐ€์ง€ ๋ฐฉ์‹์ด ์žˆ๋Š”๋ฐ ๋ฐ”๋กœ Microaveraging ๊ณผ Macroaveraging์ด๋‹ค. ๋‘˜์˜ ์ฐจ์ด๋ฅผ ํ™•์ธํ•ด๋ณด์ž. Macroaveraging ๊ตฌํ•˜๊ณ  ์‹ถ์€ ๊ฐ๊ฐ์˜ ํด๋ž˜์Šค์— ๋Œ€ํ•ด์„œ ์„ฑ๋Šฅ์„ ๊ณ„์‚ฐํ•œ๋‹ค. ๊ทธ ์ดํ›„ ๊ฐ ํด๋ž˜์Šค์— ๋Œ€ํ•œ ํ‰๊ท ๊ฐ’์ด ์ตœ์ข… ์„ฑ๋Šฅ์ด ๋œ๋‹ค. ๋งŒ์•ฝ ํด๋ž˜์Šค๊ฐ€ 4๊ฐœ๋ผ๋ฉด, ์ž‘์€ confusion matrix 4๊ฐœ๋ฅผ ๊ทธ๋ฆฌ๊ณ , ์ด 4๊ฐœ์— ๋Œ€ํ•œ ํ‰๊ท ๊ฐ’..
list_img
Ch01. Speaking Mathematically
user-img iamnotwhale
2023.09.08
Statement = Proposition : ๋ช…์ œ - ์ •์˜: ์ฐธ์ด๋‚˜ ๊ฑฐ์ง“์œผ๋กœ ํŒ๋‹จํ•  ์ˆ˜ ์žˆ๋Š” ๋ฌธ์žฅ - ์ข…๋ฅ˜ 1) Universal Statement: ํ•œ ์ง‘ํ•ฉ์˜ ๋ชจ๋“  ์š”์†Œ์— ๋Œ€ํ•ด์„œ ์ฐธ 2) Conditional Statement: ํŠน์ • ์กฐ๊ฑด๋งŒ ํฌํ•จ 3) Existential Statement: ์กฐ๊ฑด์„ ๋งŒ์กฑํ•˜๋Š” ์š”์†Œ๊ฐ€ 1๊ฐœ ์ด์ƒ ์žˆ๋‹ค. 4) Universal Conditional Statement: universal & conditional 5) Universal Existential Statement: ์ฒซ ๋ถ€๋ถ„์€ universal, ๋‘ ๋ฒˆ์งธ ๋ถ€๋ถ„์€ existential 6) Existential Universal Statement: ์ฒซ ๋ถ€๋ถ„์€ existential, ๋‘ ๋ฒˆ์งธ ๋ถ€๋ถ„์€ universal S..
list_img
[DEG ๋ถ„์„] Volcano Plot, MA Plot
user-img iamnotwhale
2023.08.10
์ตœ๊ทผ์— ๋ฐ”์ด์˜ค์ธํฌ ์ชฝ ๋žฉ์—์„œ ํ•™๋ถ€ ์—ฐ๊ตฌ์ƒ์„ ํ•˜๊ฒŒ ๋˜์—ˆ๋Š”๋ฐ, ์ƒ๊ฐ๋ณด๋‹ค ํ•œ๊ตญ์–ด ์ž๋ฃŒ๊ฐ€ ๋งŽ์ง€ ์•Š์•„์„œ ๋‚จ๊ฒจ๋ณด๋Š” ๊ธฐ๋ก! ํ‹€๋ฆฐ ์„ค๋ช…์ด ์žˆ์„ ์ˆ˜๋„ ์žˆ์Œ ์ฃผ์˜ DEG ๋ถ„์„์ด๋ž€? DEG๋Š” Differentially Expressed Gene์˜ ์•ฝ์ž! ์œ ์ „์ž์˜ ๋ฐœํ˜„๊ฐ’์„ ์ธก์ •ํ•˜๊ณ  ํ†ต๊ณ„์  ๊ณผ์ •์„ ๊ฑฐ์ณ ๋Œ€์กฐ๊ตฐ๊ณผ ๋น„๊ต๊ตฐ ๊ฐ„์— ๋ฐœํ˜„์ด ์œ ์˜ํ•œ ์œ ์ „์ž ํ›„๋ณด๊ตฐ์„ ์„ ์ •ํ•˜๋Š” ๋ฐฉ์‹์ด๋‹ค. (์ถœ์ฒ˜) DEG ๋ถ„์„์„ ์œ„ํ•ด์„œ๋Š” RNA-seq ๋ฐ์ดํ„ฐ๊ฐ€ ํ•„์š”ํ•˜๋‹ค. DEG ๋ถ„์„์„ ์œ„ํ•œ RNA-seq ๋ฐ์ดํ„ฐ๋ฅผ ๋งŒ๋“œ๋Š” ๋ฐฉ๋ฒ•์€ ์—ฌ๋Ÿฌ ๊ฐ€์ง€๊ฐ€ ์žˆ๊ฒ ์ง€๋งŒ ๋‚ด๊ฐ€ ์‚ฌ์šฉํ•œ ๋ฐฉ๋ฒ•์€ ๋‹ค์Œ๊ณผ ๊ฐ™๋‹ค. 1) RNA-seq๋ฅผ ํ†ตํ•ด ์–ป์€ fq ํŒŒ์ผ์—์„œ ํ•„์š” ์—†๋Š” adapter ๋“ฑ์˜ ์„œ์—ด์„ ์ž˜๋ผ๋ƒ„. 2) TopHat2๋ฅผ ์ด์šฉํ•œ mapping ์ง„ํ–‰. 3) Cufflinks, Cuffdif..
list_img
Categorical Data Encoding: Mean Target Encoding
user-img iamnotwhale
2023.05.31
์ด๋ฒˆ ํฌ์ŠคํŒ…์—์„œ๋Š” ์นดํ…Œ๊ณ ๋ฆฌ (๋ฒ”์ฃผํ˜•) ์ž๋ฃŒ ์ธ์ฝ”๋”ฉ ๋ฐฉ๋ฒ• ์ค‘ ํ•˜๋‚˜์ธ Mean Target Encoding์— ๋Œ€ํ•ด ์†Œ๊ฐœํ•˜๊ณ ์ž ํ•œ๋‹ค. ์‚ฌ์‹ค ๋‚˜๋Š” ๋ฒ”์ฃผํ˜• ๋ณ€์ˆ˜ ์ธ์ฝ”๋”ฉ ๋ฐฉ๋ฒ•์— ๋ผ๋ฒจ์ธ์ฝ”๋”ฉ๊ณผ ์› ํ•ซ ์ธ์ฝ”๋”ฉ๋งŒ ์žˆ๋Š” ์ค„ ์•Œ์•˜๋‹ค. ๊ทธ๋Ÿฐ๋ฐ ์ด๋ฒˆ ์ฑ„๋ฌด๋ถˆ์ดํ–‰ ๋ฐ์ดํ„ฐ๋ฅผ ๋ถ„์„ํ•  ๋•Œ ๋‚ด๊ฐ€ ํ–ˆ๋˜ ์ ˆ์ฐจ๋“ค์ด ์‚ฌ์‹ค์€ Mean Target Encoding์ด์—ˆ๋‹ค๋Š” ๊ฒƒ์„ ์•Œ๊ฒŒ ๋˜์–ด ๋†€๋ž„ ์ˆ˜๋ฐ–์— ์—†์—ˆ๋‹ค. ์•„๋ฌดํŠผ ๋ณธ๋ก ์œผ๋กœ ๋“ค์–ด๊ฐ€์ž. ๋ฒ”์ฃผํ˜• ๋ณ€์ˆ˜๊ฐ€ ๋ฌด์—‡์ธ์ง€๋Š”, ์ด์ „์— ๋‚ด๊ฐ€ ์ž‘์„ฑํ•˜์˜€๋˜ ๊ธ€์„ ์ฐธ๊ณ ํ•˜๋ฉด ์ข‹์„ ๊ฒƒ ๊ฐ™๋‹ค. https://iamnotwhale.tistory.com/5 [EDA] ๋ฐ์ดํ„ฐ ์ข…๋ฅ˜๋ณ„ ์‹œ๊ฐํ™” ๋ฐฉ๋ฒ• ๋ฐ์ดํ„ฐ์˜ ์ข…๋ฅ˜ ๋ฐ์ดํ„ฐ๋Š” ํฌ๊ฒŒ ๋ฒ”์ฃผํ˜•, ์ˆ˜์น˜ํ˜• ๋‘ ๊ฐ€์ง€ ๋ถ„๋ฅ˜๋กœ ๋‚˜๋ˆŒ ์ˆ˜ ์žˆ๋‹ค. ๋ฒ”์ฃผํ˜• ๋ฐ์ดํ„ฐ๋Š” ๋ฒ”์ฃผ/์นดํ…Œ๊ณ ๋ฆฌ๋ฅผ ๊ตฌ๋ถ„ํ•˜๋Š” ๊ฐ๊ฐ์˜ ์ด๋ฆ„์„ ๊ฐ–๋Š” ๋ฐ์ด..
Chapter 3. Informed Search
user-img iamnotwhale
2023.05.24
Greedy Best-First Search Evaluation Func & Heuristic Func Evaluation func f(n) node n์˜ estimated cost Heuristic func h(n) node n์—์„œ goal state๋กœ ๊ฐ€๋Š” ๊ฐ€์žฅ ์ €๋ ดํ•œ ๋น„์šฉ์˜ ์ถ”์ •์น˜ root-finding problem์—์„œ ๊ฐ ๋„์‹œ๋ณ„ ์ง์„ ๊ฑฐ๋ฆฌ ํ•จ์ˆ˜ Greedy Best-First Search ์ •์˜: best-first search์˜ ์ผ์ข… → h(n) value๊ฐ€ ๊ฐ€์žฅ ์ ์€ ๋…ธ๋“œ๋ฅผ ํ™•์žฅ table์„ ์œ ์ง€ํ•˜๋Š” ๊ฒฝ์šฐ ๊ทธ๋ž˜ํ”„ (์ค‘๋ณต์„ ๊ณ ๋ ค) ์—†์• ๋Š” ๊ฒฝ์šฐ tree search (์ค‘๋ณต์„ ๊ณ ๋ คํ•˜์ง€ ์•Š์Œ) ์„ฑ๋Šฅ ํ‰๊ฐ€ Completeness → ์‚ฌ์ดํด ๋ฐœ์ƒ Not complete: tree-like search co..
1 2 3 4 5 6