๊ด€๋ฆฌ ๋ฉ”๋‰ด

๋ชฉ๋ก2016/07/04 (1)

Wookang makes AI

ํ˜„๋Œ€๊ฒฝ์ œ์—ฐ๊ตฌ์› ๋ณด๊ณ ์„œ Scikit-learn๊ณผ scipy๋ฅผ ์ด์šฉํ•œ ๋น„๊ณ„์ธต ๊ตฐ์ง‘ ๋ถ„์„

โ— ํ˜„๋Œ€๊ฒฝ์ œ์—ฐ๊ตฌ์› ๋ณด๊ณ ์„œ Scikit-learn๊ณผ scipy๋ฅผ ์ด์šฉํ•œ ๋น„๊ณ„์ธต ๊ตฐ์ง‘ ๋ถ„์„ > ๋“ค์–ด๊ฐ€๋Š” ๋ง 1. ์ด์ œ (์•ฝ๊ฐ„์˜)์ธ๊ณต์ง€๋Šฅ์ด ๋“ค์–ด๊ฐ„๋‹ค. ์•Œ์•„์„œ ๋ฌธ์„œ์˜ ์ค‘์‹ฌ(centroid)์„ ์„ค์ •ํ•˜๊ณ  ์ด๋กœ๋ถ€ํ„ฐ ๊ฐ ์ž๋ฃŒ์™€์˜ ๊ฑฐ๋ฆฌ์— ๋“œ๋Š” ๋น„์šฉ์„ ์ตœ์†Œํ•œํ•˜๋Š” ๊ตฐ์ง‘ ์•Œ๊ณ ๋ฆฌ์ฆ˜(K-ํ‰๊ท ๊ธฐ๋ฒ•)์„ ์ด์šฉํ•œ๋‹ค. ์ด๊ฒƒ๋„ ๊ฒฐ๊ณผ๊ฐ€ ๊ถ๊ธˆํ•˜๋‹ค. > ๊ณ„ํš1. scikit-learn ๋ผ์ด๋ธŒ๋Ÿฌ๋ฆฌ์—์„œ ์ด๋ฏธ K-ํ‰๊ท  ๊ตฐ์ง‘ํ™” ๊ธฐ๋Šฅ์„ ์ œ๊ณตํ•œ๋‹จ๋‹ค.2. ์ฒ˜๋ฆฌ ์†๋„๊ฐ€ ๋นจ๋ผ ๋งŽ์€ ๋ถ„์•ผ์—์„œ ์‚ฌ์šฉํ•˜๊ณ  ์žˆ๋‹จ๋‹ค.3. ๋ฌธ์„œ๋ฅผ ํ–‰๋ ฌ๋กœ ๋งŒ๋“ค๋•Œ ๊ฐœ๋ฐœ์ž๊ฐ€ ์ผ์ผ์ด ๋งŒ๋“ค ํ•„์š”์—†์ด vectorizer = TfidfVectorizer(min_df=1) doc_term_mat = vectorizer.fit_transform(documents) ์ด๋ ‡๊ฒŒ ํ•˜๋Š” ๊ฒƒ์œผ๋กœ ์›์ƒท ์ƒ์„ฑ์ด ๊ฐ€๋Šฅํ•˜๋‹ค..

๊ทธ ๋ฐ–์— AI 2016. 7. 4. 19:09