[B! scikit-learn] johshishaã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

johshisha id:johshisha

scikit-learnã«é–¢ã™ã‚‹johshishaã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (7)

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

${{author_name}}$
{{author_name}}{{created}}
{{ #comment }}{{ comment }}{{ /comment }}
- {{ label }}

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

scikit-learnã¨gensimã§ãƒ‹ãƒ¥ãƒ¼ã‚¹è¨˜äº‹ã‚’åˆ†é¡žã™ã‚‹ - Qiita
ã“ã‚“ã«ã¡ã¯ã€åˆå¿ƒè€…ã§ã™ã€‚ é©å½“ãªãƒ‹ãƒ¥ãƒ¼ã‚¹è¨˜äº‹ãŒã‚ã£ãŸã¨ã—ã¦ã€ãƒ‹ãƒ¥ãƒ¼ã‚¹ã®ã‚«ãƒ†ã‚´ãƒªã‚’æŽ¨æ¸¬ã™ã‚‹ã¿ãŸã„ãªã€ã‚ˆãã‚ã‚‹ã‚„ã¤ã‚’ã‚„ã£ã¦ã¿ã¾ã—ãŸã€‚Python3.3ã‚’ä½¿ã„ã¾ã—ãŸã€‚ ä½•ã‚’ã‚„ã‚‹ã®ï¼Ÿ ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã¯livedoorãƒ‹ãƒ¥ãƒ¼ã‚¹ã‚³ãƒ¼ãƒ‘ã‚¹ã‚’ä½¿ã„ã¾ã—ãŸã€‚ http://www.rondhuit.com/download.html#ldcc ã‚¯ãƒªã‚¨ã‚¤ãƒ†ã‚£ãƒ–ãƒ»ã‚³ãƒ¢ãƒ³ã‚ºãƒ©ã‚¤ã‚»ãƒ³ã‚¹ãŒé©ç”¨ã•ã‚Œã‚‹ãƒ‹ãƒ¥ãƒ¼ã‚¹è¨˜äº‹ã ã‘ã‚’é›†ã‚ã¦ã‚‹ãã†ã§ã™ã€‚ ãƒˆãƒ”ãƒƒã‚¯ãƒ‹ãƒ¥ãƒ¼ã‚¹ã€Sports Watchã€ITãƒ©ã‚¤ãƒ•ãƒãƒƒã‚¯ã€å®¶é›»ãƒãƒ£ãƒ³ãƒãƒ« ã€MOVIE ENTERã€ç‹¬å¥³é€šä¿¡ã€ã‚¨ã‚¹ãƒžãƒƒã‚¯ã‚¹ã€livedoor HOMMEã€Peachy ã¨ã„ã†ã‚¯ãƒ©ã‚¹ãŒã‚ã‚Šã¾ã™ã€‚ ãƒ‡ãƒ¼ã‚¿ã¯ã€1è¨˜äº‹1ãƒ†ã‚ã‚¹ãƒˆãƒ•ã‚¡ã‚¤ãƒ«ã®å½¢å¼ã§ã€ã‚¯ãƒ©ã‚¹åˆ¥ã®ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã«ã„ã£ã±ã„å…¥ã£ã¦ã„ã¾ã™ã€‚ ã“ã‚Œã‚’å¦ç¿’ã—ã¦ã€æœªçŸ¥ã®æ–‡ç« ã«å¯¾ã—ã¦ã€ãŠå‰ã¯ç‹¬å¥³é€šä¿¡ã£ã½ã„ã€ãŠå‰ã¯å®¶é›»ãƒãƒ£ãƒ³ãƒãƒ«ã£ã½ã„ã€ã¿ãŸã„ã«ã€åˆ†é¡žãŒ
johshisha 2016/11/28
python

mecab

scikit-learn

bow

å½¢æ…‹ç´ è§£æž
ãƒªãƒ³ã‚¯
Python pandas ã®ãƒ‡ãƒ¼ã‚¿ã‚’ scikit-learn ã§ã†ã¾ã„ã“ã¨å‡¦ç†ã—ãŸã„ - StatsFragments
ã¯ã˜ã‚ã« Python ã§æ©Ÿæ¢°å¦ç¿’ã™ã‚‹å ´åˆã€pandas ã§å‰å‡¦ç†ã—ãŸãƒ‡ãƒ¼ã‚¿ã‚’ scikit-learn ã§å‡¦ç†ã™ã‚‹ã€ã¨ã„ã†ã‚±ãƒ¼ã‚¹ãŒå¤šã„ã¨æ€ã†ã€‚pandas, scikit-learn ã«ã¯ ãã‚Œãžã‚Œ ç°¡å˜ã«ã§ãã‚‹ / ã§ããªã„å‡¦ç†ãŒã‚ã‚‹ã®ã§ã€ã†ã¾ã„ã“ã¨é€£æºã§ãã‚‹ã¨ã†ã‚Œã—ã„ã€‚ scikit-learn ã®å„ãƒ¡ã‚½ãƒƒãƒ‰ã¯ numpy.ndarray ã«å¯¾ã™ã‚‹å‡¦ç†ã‚’å‰æã«ã—ã¦ã„ã‚‹ãŸã‚ã€pandas ã®ãƒ‡ãƒ¼ã‚¿å½¢å¼ (DataFrame ã‚„ Series) ã‚’æ¸¡ã™ã¨ å†…éƒ¨ã§ ndarray ã«å¤‰æ›ã—ã¦çµæžœã‚’è¿”ã—ã¦ãã‚‹ã€‚ãã®ãŸã‚ã€çµæžœã«å¯¾ã—ã¦ ç›´æŽ¥ pandas ã®å‡¦ç†ã‚’ç¶šã‘ã‚‹ã“ã¨ã¯ã§ããªã„ã€‚ ndarray ã§å‡¦ç†ã™ã‚Šã‚ƒã„ã„ã˜ã‚ƒã‚“ã€ã€ã¨ã„ã†ã‚€ãã‚‚ã‚ã‚‹ãŒã€è‡ªåˆ†ã¯ã©ã†ã—ã¦ã‚‚ pandas ã§å‡¦ç†ãŒã—ãŸã„ã‚“ã ã€‚ã¨ã‚Šã‚ãˆãšã€pandas ã®ãƒ‡ãƒ¼ã‚¿ã‚’ã§ãã‚‹ã ã‘ç¶æŒã—ãŸã„ã€ã¨ã„ã†ãƒ¢ãƒãƒ™ãƒ¼ã‚·ãƒ§ãƒ³ãŒã‚ã‚‹ã‚‚ã®ã¨ã—ã¦å‡¦
johshisha 2016/10/05
scikit-learn

pandas
ãƒªãƒ³ã‚¯
Machine Learning with Scikit Learn (Part III) - äººå·¥çŸ¥èƒ½ã«é–¢ã™ã‚‹æ–å‰µéŒ²
Machine Learning with Scikit Learn (Part II)ï¼ˆ2015/8/26ï¼‰ã®ã¤ã¥ãã€‚ä»Šå›žã¯ã€Part IIã®å‹•ç”»ã®5.1ç¯€ã®å†…å®¹ã‚’ç°¡å˜ã«ã¾ã¨ã‚ãŸã€‚æ›¸ã„ã¦ãŸã‚‰é•·ããªã£ãŸã®ã§5.1ç¯€ã ã‘ã€‚ 5.1 In Depth - Linear Models å›žå¸°ã®ãŸã‚ã®ç·šå½¢ãƒ¢ãƒ‡ãƒ«ã®è©³ç´°ãŒå–ã‚Šä¸Šã’ã‚‰ã‚Œã¦ã„ã‚‹ã€‚ç·šå½¢å›žå¸°ã¯éŽå¦ç¿’ï¼ˆOverfittingï¼‰ã«å¼±ã„ã€‚éŽå¦ç¿’ã‚’å›žé¿ã™ã‚‹ãŸã‚ã«æ£å‰‡åŒ–ã‚’å°Žå…¥ã—ãŸãƒªãƒƒã‚¸å›žå¸°ã€Lassoã€ElasticNetã®ä½¿ã„æ–¹ãŒäº‹ä¾‹ã‚’ã‚‚ã¨ã«èª¬æ˜Žã•ã‚Œã¦ã„ã‚‹ã€‚ã¾ãŸã€ç·šå½¢åˆ†é¡žå™¨ã®æ£å‰‡åŒ–ã®ä¾‹ã¨ã—ã¦SVMã¨ãƒã‚¸ã‚¹ãƒ†ã‚£ãƒƒã‚¯å›žå¸°ã®ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿CãŒèª¬æ˜Žã•ã‚Œã¦ã„ã‚‹ã€‚ ç·šå½¢å›žå¸° ã¾ãšã¯ã€å˜ç´”ãªç·šå½¢å›žå¸°ã‹ã‚‰ã€‚ç·šå½¢å›žå¸°ã¯ã€ç‰¹å¾´é‡ã®ç·šå½¢çµåˆã§è¡¨ã•ã‚Œã‚‹ãƒ¢ãƒ‡ãƒ«ã€‚ ã“ã“ã§ã€ãŒä¿‚æ•°ï¼ˆcoef_ï¼‰ã§ãŒåˆ‡ç‰‡ï¼ˆintercept_ï¼‰ã¨å‘¼ã°ã‚Œã‚‹ã€‚ç·šå½¢å›žå¸°ã§ã¯ã€æ•™å¸«ãƒ‡ãƒ¼ã‚¿ ã‹ã‚‰ä¿‚æ•°ã¨åˆ‡ç‰‡ã‚’å¦ç¿’ã™
johshisha 2016/09/21
scikit-learn
ãƒªãƒ³ã‚¯
Pythonã§Bag of Wordsã¨SVMã‚’ä½¿ã£ãŸã‚¿ã‚¤ãƒˆãƒ«ã®ã‚«ãƒ†ã‚´ãƒªåˆ†é¡ž - stMind
cc licensed ( BY ) flickr photo shared by Loco Steve é€±æœ«ã«è©¦ãã†ã®ã‚³ãƒ¼ãƒŠãƒ¼ã€‚ ã¡ã‚‡ã†ã©è‰¯ã„ãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«ãŒã‚ã£ãŸã®ã§ã€ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã‚’ç”¨æ„ã—ã¦ã‚„ã£ã¦ã¿ã¾ã—ãŸã€‚ å•é¡Œ How can I get a computer to tell me what an article is about (provided methods such as bribery and asking politely do not work)? ã‚ã‚‹è¨˜äº‹ãŒä½•ã«ã¤ã„ã¦æ›¸ã‹ã‚Œã¦ã„ã‚‹ã®ã‹ã€ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ã«ç†è§£ã•ã›ã‚‹ã«ã¯ã©ã†ã™ã‚Œã°è‰¯ã„ã‹? ãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«ã§ã¯æ‰‹å‹•ã§ä½œã£ãŸãƒ‡ãƒ¼ã‚¿ã‚’ä½¿ã£ã¦çŠ¬ã‚‚ã—ãã¯ã‚µãƒ³ãƒ‰ã‚¤ãƒƒãƒã®2ã‚¯ãƒ©ã‚¹åˆ†é¡žã‚’ã—ã¦ã„ã¾ã™ã€‚ ã“ã“ã§ã¯ã€Google Newsã§iPadã®ãƒ‹ãƒ¥ãƒ¼ã‚¹ã€ã‚½ãƒäº”è¼ªã®ãƒ‹ãƒ¥ãƒ¼ã‚¹ã¨ã‚«ãƒ†ã‚´ãƒªåˆ†ã‘ã•ã‚Œã¦ã„ã‚‹è¨˜äº‹ã®ã‚¿ã‚¤ãƒˆãƒ«ã‚’ä½¿ã£ã¦ã€ ã‚ã‚‹ã‚¿ã‚¤ãƒˆãƒ«ãŒiPa
johshisha 2016/08/18
python

nlp

scikit-learn
ãƒªãƒ³ã‚¯
scikit-learnã§æ©Ÿæ¢°å¦ç¿’ã‚’è©¦ã™ SVM - Qiita
ã“ã‚“ã«ã¡ã¯ã€‚åˆå¿ƒè€…ã§ã™ã€‚ ã‚ã‚‹ã‚µãƒ¼ãƒ“ã‚¹ï¼ˆå¿µå†™ã§ãã‚‹äººã¨ä½•ã‹å¿µå†™ã—ã¦ã»ã—ã„äººã®ãƒžãƒƒãƒãƒ³ã‚°ã‚µãƒ¼ãƒ“ã‚¹ã§ã™ï¼‰ã§æ©Ÿæ¢°å¦ç¿’ãŸã‚ãã†ã¨æ€ã£ã¦ã„ã¦ã€èª¿ã¹ã¦ã„ã‚‹æœ€ä¸ã§ã™ã€‚ å•é¡Œ è©¦ã—ã«ã€ã“ã‚Œã®1å•ç›®ã‚’è§£ãã‚ˆãƒ¼ã€‚ æ•™å¸«ã‚ã‚Šå¦ç¿’ã®å•é¡Œã§ã™ã€‚ä¸Šè¨˜ãƒšãƒ¼ã‚¸ã®å›žç”ä¾‹ã®ã‚ˆã†ã«ã€SVMã‚’ä½¿ã£ã¦ã¿ã¾ã™ã€‚ ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã¯ã€scikit-learnã‚’ä½¿ã£ã¦ã¿ãŸã€‚ SVMä½¿ã„ãŸã„ã ã‘ãªã‚‰ã»ã‹ã«ã‚‚ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã‚ã‚‹ã¿ãŸã„ã ã‘ã©ã€scikit-learnã„ã„ã‚ˆ~çš„ãªã“ã¨ã‚’è¨€ã†äººãŒå¤šã„ã®ã§ä½¿ã£ã¦ã¿ã¾ã—ãŸã€‚ ã¨ã‚Šã‚ãˆãšã‚„ã£ã¦ã¿ã‚‹ # -*- coding: utf-8 -*- from sklearn.svm import LinearSVC import numpy as np # å¦ç¿’ãƒ‡ãƒ¼ã‚¿ data_training_tmp = np.loadtxt('CodeIQ_auth.txt', delimiter=' ') data_train
johshisha 2016/03/09
scikit-learn

SVM
ãƒªãƒ³ã‚¯
scikit-learnã®SVMï¼ˆSVCï¼‰ã®å‡¦ç†é€Ÿåº¦ã«ã¤ã„ã¦ - Qiita
2016.09.14 å‡¦ç†æ™‚é–“ã®ãƒãƒ©ã¤ãã«ã¤ã„ã¦è¿½è¨˜ã—ã¾ã—ãŸ scikit-learnã®SVCï¼ˆrbfã‚«ãƒ¼ãƒãƒ«ã¨linearã‚«ãƒ¼ãƒãƒ«ï¼‰ã¨LinearSVCã®å‡¦ç†é€Ÿåº¦ã‚’æ¯”è¼ƒã—ã¦ã¿ã¾ã—ãŸï¼Ž åˆ©ç”¨ã—ãŸãƒ‡ãƒ¼ã‚¿ã¯Rã®kernlabãƒ‘ãƒƒã‚±ãƒ¼ã‚¸ã«å«ã¾ã‚Œã¦ã„ã‚‹spamãƒ‡ãƒ¼ã‚¿ã§ã™ï¼Ž èª¬æ˜Žå¤‰æ•°ã¯4601ã‚µãƒ³ãƒ—ãƒ«ï¼Œ57æ¬¡å…ƒï¼Œ ãƒ©ãƒ™ãƒ«ã¯spam:1813ã‚µãƒ³ãƒ—ãƒ«ï¼Œnonspam:2788ã‚µãƒ³ãƒ—ãƒ«ã§ã™ï¼Ž ã‚µãƒ³ãƒ—ãƒ«æ•°ï¼Œæ¬¡å…ƒæ•°ã‚’å¤‰ãˆãŸæ™‚ã®çµæžœã¯ä»¥ä¸‹ã®é€šã‚Šã§ã™ï¼Ž SVCã®linearã‚«ãƒ¼ãƒãƒ«ãŒé…ã™ãŽã¾ã™ãï¼Ž ã¤ã„ã¤ã„ã‚«ãƒ¼ãƒãƒ«ç¨®åˆ¥ã¾ã§å«ã‚ã¦ã‚°ãƒªãƒƒãƒ‰ã‚µãƒ¼ãƒã—ã¦ã—ã¾ã„ãŸããªã‚Šã¾ã™ãŒï¼Œ ãã¡ã‚“ã¨LinearSVCã‚’ä½¿ã£ãŸã»ã†ãŒè‰¯ã•ãã†ã§ã™ï¼Ž æ¤œè¨¼ç”¨ã‚³ãƒ¼ãƒ‰ã¯ä»¥ä¸‹ï¼Ž å‡¦ç†æ™‚é–“è¨ˆæ¸¬ã®éƒ½åˆã§ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿Cã‚’æŒ¯ã£ã¦ã„ã¾ã™ï¼Ž ã¾ãŸç‰¹å¾´é‡é¸æŠžï¼ˆæ¬¡å…ƒå‰Šæ¸›ï¼‰ã¯RandomForestã®feature importanceã‚’åˆ©ç”¨ã—ã¾ã—ãŸï¼Ž ã“ã‚Œã¯é©å½“
johshisha 2016/03/09
scikit-learn

SVM
ãƒªãƒ³ã‚¯
pythonã®æ©Ÿæ¢°å¦ç¿’ãƒ©ã‚¤ãƒ–ãƒ©ãƒªscikit-learnã®ç´¹ä»‹ - å”¯ç‰©æ˜¯çœŸ @Scaled_Wurm
scikit-learn(sklearn)ã®æ—¥æœ¬èªžã®å…¥é–€è¨˜äº‹ãŒã‚ã‚“ã¾ã‚Šãªã„ãªãƒ¼ã¨æ€ã£ã¦æ›¸ãã¾ã—ãŸã€‚ ã©ã¡ã‚‰ã‹ã£ã¦ã„ã†ã¨ã‚ˆãä½¿ã†æ©Ÿèƒ½ã®ç´¹ä»‹çš„ãªæ„Ÿã˜ã§ã™ã€‚ è‹±èªžãŒèªã‚ã‚‹æ–¹ã¯å…¬å¼ã®ãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«ãŒãŠã™ã™ã‚ã§ã™ã€‚ scikit-learnã¨ã¯ï¼Ÿ scikit-learnã¯ã‚ªãƒ¼ãƒ—ãƒ³ã‚½ãƒ¼ã‚¹ã®æ©Ÿæ¢°å¦ç¿’ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã§ã€åˆ†é¡žã‚„å›žå¸°ã€ã‚¯ãƒ©ã‚¹ã‚¿ãƒªãƒ³ã‚°ãªã©ã®æ©Ÿèƒ½ãŒå®Ÿè£…ã•ã‚Œã¦ã„ã¾ã™ã€‚ ã¾ãŸæ§˜ã€…ãªè©•ä¾¡å°ºåº¦ã‚„ã‚¯ãƒã‚¹ãƒãƒªãƒ‡ãƒ¼ã‚·ãƒ§ãƒ³ã€ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã®ã‚°ãƒªãƒƒãƒ‰ã‚µãƒ¼ãƒãªã©ã®ç—’ã„ã¨ã“ã‚ã«æ‰‹ãŒå±Šãæ©Ÿèƒ½ã‚‚ã‚ã‚Šã¾ã™ã€‚ ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ« scikit-learnã®ä»–ã«ã‚‚numpyã¨ã‹scipyã¨ã‹ãŒå¿…è¦ã§ã™ã€‚ Windows 64 bitç‰ˆã®äººã¯ä»¥ä¸‹ã®URLã«è‰²ã€…ãªã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ©ãƒ¼ãŒãŠã„ã¦ã‚ã‚‹ã®ã§ãŠã™ã™ã‚ Python Extension Packages for Windows - Christoph Gohlke ãã®ä»–ã®äººã¯ä»¥ä¸‹ã®URLã‚’è¦‹ã¦
johshisha 2015/11/25
python

scikit-learn
ãƒªãƒ³ã‚¯
1