Machine Learning Cheat Sheet (for scikit-learn)
As you hopefully have heard, we at scikit-learn are doing a user survey (which is still open by the way). One of the requests there was to provide some sort of flow chart on how to do machine learning. As this is clearly impossible, I went to work straight away. This is the result: [edit2] clarification: With ensemble classifiers and ensemble regressors I mean random forests, extremely randomized
Brains, Sex, and Machine Learning
Geoffrey Hinton University of Toronto Abstract: Recent advances in machine learning cast new light on two puzzling biological phenomena. Neurons can use the precise time of a spike to communicate a real value very accurately, but it appears that cortical neurons do not do this. Instead they send single, randomly timed spikes. This seems like a clumsy way to perform signal processing, but a recent
http://www.simplex.t.u-tokyo.ac.jp/~s-taiji/ibisml2012/IBIS2012.pdf
. ...... 統計的å¦ç¿’ç†è«–ãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«: 基礎ã‹ã‚‰å¿œç”¨ã¾ã§ †鈴木 大慈 †æ±äº¬å¤§å¦ æƒ…å ±ç†å·¥å¦ç ”究科 æ•°ç†æƒ…å ±å¦å°‚æ”» IBIS 2012@ç‘波大å¦æ±äº¬ã‚ãƒ£ãƒ³ãƒ‘ã‚¹æ–‡äº¬æ ¡èˆŽ 2012 å¹´ 11 月 7 æ—¥ 1 / 60 æ§‹æˆ ...1 ã¯ã˜ã‚ã«: ç†è«–ã®å½¹å‰² ...2 統計的å¦ç¿’ç†è«–ã¨çµŒé¨“éŽç¨‹ ...3 一様ãƒã‚¦ãƒ³ãƒ‰ 基本的ãªä¸ç‰å¼ Rademacher 複雑ã•ã¨ Dudley ç©åˆ† 局所 Rademacher 複雑㕠...4 最é©æ€§ 許容性 minimax 最é©æ€§ ...5 ベイズã®å¦ç¿’ç†è«– 2 / 60 æ§‹æˆ ...1 ã¯ã˜ã‚ã«: ç†è«–ã®å½¹å‰² ...2 統計的å¦ç¿’ç†è«–ã¨çµŒé¨“éŽç¨‹ ...3 一様ãƒã‚¦ãƒ³ãƒ‰ 基本的ãªä¸ç‰å¼ Rademacher 複雑ã•ã¨ Dudley ç©åˆ† 局所 Rademacher 複雑㕠...4 最é©æ€§ 許容性 minimax 最é©æ€§ ...5 ベイズã®
Installing BLAS/Lapack/ITPP on Amazon EC2/Ubuntu Linux
BLAS/Lapack are efficient matrix math libraries. The following instructions explains how to install them for Amazon EC2 (Ubuntu maverick version, and Amazon Linux). It++ (itpp) is a popular c++ wrapper for blas/lapack. DISLAIMER: The below instructions are for 64 bit machines. For 32 bit machines follow other instructions: http://bickson.blogspot.com/2011/06/graphlab-pmf-on-32-bit-linux.html FOR L
é‡åアニーリング解説 1
2. アウトライン •  é‡å力å¦å…¥é–€ –  状態ã€ç´”粋状態ã€æ··åˆçŠ¶æ…‹ •  å¤å…¸åŠ›å¦ã®ä¸–ç•Œã¨é‡å力å¦ã®ä¸–ç•Œ •  é‡å力å¦ã®ä¸–ç•Œ –  純粋状態ã¨æ··åˆçŠ¶æ…‹ •  機械å¦ç¿’ã¸ã®å¿œç”¨ –  クラス分類å•é¡Œ •  é‡åアニーリングã«ã‚ˆã‚‹æœ€é©åŒ– •  to be con1nued 4. 状態ã€ç´”粋状態ã€æ··åˆçŠ¶æ…‹ •  å¤å…¸åŠ›å¦ã®ä¸–ç•Œ ç²’åã®çŠ¶æ…‹ (x, p) : 相空間ã®åº§æ¨™ ä½ç½®ã¨é‹å‹•é‡ã‚’決ã‚れ㰠å¤å…¸çš„ãªç²’åã®çŠ¶æ…‹ã¯å®Œå…¨ã«æ±ºã¾ã‚‹ •  é‡å力å¦ã®ä¸–ç•Œ ç²’åã®çŠ¶æ…‹ | i ： 状態ベクトル ä½ç½®ã¨é‹å‹•é‡ã‚’åŒæ™‚ã«æ±ºã‚ã‚‹ã“ã¨ãŒã§ããªã„ （ä¸ç¢ºå®šæ€§åŽŸç†ï¼‰ 物ç†çš„ã«å¯èƒ½ãªé™ã‚Šæƒ…å ±ã‚’æŒ‡å®šã—å°½ãã—ãŸã¨ã 「指定ã®ä»•æ–¹ã€ãŒã€ã‚るベクトル空間ã®å…ƒã¨ãªã‚‹ ï¼ˆè¤‡ç´ ãƒ’ãƒ«ãƒ™ãƒ«ãƒˆç©ºé–“ï¼‰ 5. 状態ã€ç´”粋状態ã€æ··åˆçŠ¶æ…‹ •  純粋状態 –  å‰è¿°ã®ã€Œç‰©ç†çš„ã«å¯èƒ½ãªé™ã‚Šæƒ…å ±ã‚’æŒ‡å®šã—å°½ãã—ãŸçŠ¶æ…‹ã€ を純粋状態ã¨å‘¼ã¶ –  æ™®
Accepted Papers | ICDM 2012
Acceptance Rate: Regular (10.7%), Short (9.27%), Total (19.97%) Regular Papers
Visual Geometry Group - University of Oxford
This website uses Google Analytics to help us improve the website content. This requires the use of standard Google Analytics cookies, as well as a cookie to record your response to this confirmation request. If this is OK with you, please click 'Accept cookies', otherwise you will see this notice on every page. For more information, please click here Accept cookies Carlos Arteta, Victor Lempitsky
SciKits - scikits
See the source. Created page in 1.070 seconds. Sign in Designed by Janto Dreijer. Appearance based on Sphinx and Phuzion icons.
Microsoft PowerPoint - NEC-MachineLearning1-jp.ppt
NEC 2009å¹´2月26æ—¥ 機械å¦ç¿’概論（１） 分類å•é¡Œã«ãŠã‘る能動å¦ç¿’ æ±äº¬å·¥æ¥å¤§å¦ 計算工å¦å°‚æ”» æ‰å±± å°† [email protected]itech.ac.jp http://sugiyama-www.cs.titech.ac.jp/~sugi/ 教師付ãå¦ç¿’ 目的：訓練標本 ã‹ã‚‰ï¼ŒçœŸã®é–¢æ•° ã®è¿‘似関数 を求ã‚ã‚‹ 未知ã®ãƒ†ã‚¹ãƒˆå…¥åŠ› ã§ã®å‡ºåŠ› ㌠推定ã§ãるよã†ã«ãªã‚‹ï¼ˆæ±ŽåŒ–能力） 出力 2 入力 汎化誤差 汎化誤差（関数ã®è¿‘ã•ã®å°ºåº¦ï¼‰ï¼š 3 ： 㧠を予測ã—ãŸã¨ãã®æ失 ：テスト入力点ãŒå¾“ã†ç¢ºçŽ‡å¯†åº¦é–¢æ•° (既知ã¨ä»®å®šï¼Žèƒ½å‹•å¦ç¿’ã§ã¯æœ¬è³ªçš„） 汎化誤差ã¯ï¼Œãƒ†ã‚¹ãƒˆèª¤å·®ã®æœŸå¾…値 汎化誤差を最å°ã«ã™ã‚‹å¦ç¿’çµæžœ 求ã‚ã‚‹ã“ã¨ãŒæ•™å¸«ä»˜ãå¦ç¿’ã®ç›®çš„ ã‚’ 回帰å•é¡Œã¨åˆ†é¡žå•é¡Œ 回帰å•é¡Œï¼š 通常ã¯äºŒä¹—æ失を用ã„る： 4 分類å•é¡Œï¼š 通常ã¯ï¼ï¼ï¼‘æ失を用ã„る： 能動å¦ç¿’（実験計画） 教師付
ç§ã®ãƒ–ックマーク: 確率密度比ã«åŸºã¥ã機械å¦ç¿’ (Machine Learning based on Density Ratio) - 人工知能å¦ä¼šèªŒ
 ç§ã®ãƒ–ックマーク: 確率密度比ã«åŸºã¥ã機械å¦ç¿’ (Machine Learning based on Density Ratio) æ±äº¬å·¥æ¥å¤§å¦ã€€å¤§å¦é™¢æƒ…å ±ç†å·¥å¦ç ”究科　計算工å¦å°‚æ”» å±±ç”°ã€€èª (Makoto Yamada) URL: http://sugiyama-www.cs.titech.ac.jp/~yamada ã¯ã˜ã‚㫠パターンèªè˜ã€ãƒ‰ãƒ¡ã‚¤ãƒ³é©å¿œã€å¤–れ値検出ã€å¤‰åŒ–点検出ã€æ¬¡å…ƒå‰Šæ¸›ã€å› 果推論ç‰ã®æ§˜ã€…ãªæ©Ÿæ¢°å¦ç¿’ã®å•é¡ŒãŒç¢ºçŽ‡å¯†åº¦æ¯”(確率密度関数ã®æ¯”)ã®å•é¡Œã¨ã—ã¦å®šå¼åŒ–ã§ãã‚‹ã“ã¨ã‹ã‚‰ã€è¿‘å¹´ã€ç¢ºçŽ‡å¯†åº¦æ¯”ã«åŸºã¥ã„ãŸæ©Ÿæ¢°å¦ç¿’ã®ç ”究ãŒæ©Ÿæ¢°å¦ç¿’ãŠã‚ˆã³ãƒ‡ãƒ¼ã‚¿ãƒžã‚¤ãƒ‹ãƒ³ã‚°ã®åˆ†é‡Žã«ãŠã„ã¦å¤§å¤‰æ³¨ç›®ã•ã‚Œã¦ã„る。本ブックマークã§ã¯ç¢ºçŽ‡å¯†åº¦æ¯”ã®ç ”究開発ã®ãŸã‚ã®æœ‰ç”¨ãªãƒªã‚½ãƒ¼ã‚¹ã‚’紹介ã™ã‚‹ã€‚ è§£èª¬è¨˜äº‹ç‰ å¯†åº¦æ¯”ã«åŸºã¥ã機械å¦ç¿’ã®æ–°ãŸãªã‚¢ãƒ—ãƒãƒ¼ãƒ 　統計数ç†, vol.58, no.2, pp.141-15
ICML2012èªã¿ä¼šã§ç™ºè¡¨ã—ã¾ã—㟠&& SVMã®æ€§èƒ½ã‚’ガタè½ã¡ã•ã›ã‚‹ãŸã‚ã«ã¯ - kisa12012ã®æ—¥è¨˜
本日サイボウズラボã•ã‚“ã®ä¼šå ´ã§é–‹å‚¬ã•ã‚ŒãŸICML2012èªã¿ä¼šã«ç™ºè¡¨è€…ã¨ã—ã¦å‚åŠ ã—ã¾ã—ãŸï¼Ž 主催者ã®nokunoã•ã‚“ï¼Œä¼šå ´ä¿‚ã®shuyoã•ã‚“，ã¾ãŸå‚åŠ è€…ã®çš†æ§˜ï¼Œã‚ã‚ŠãŒã¨ã†ã”ã–ã„ã¾ã—ãŸï¼éžå¸¸ã«å‹‰å¼·ã«ãªã‚Šã¾ã—ãŸï¼Ž 今回発表ã—ãŸã®ã¯ï¼ŒPoisoning Attacks against Support Vector Machines (Biggio+) ã§ã™ï¼Ž 発表資料 PoisoningAttackSVM (ICMLreading2012) from Hidekazu Oiwa 元論文 http://icml.cc/2012/papers/880.pdf æ¦‚è¦ è«–æ–‡ã®åå‰é€šã‚Šï¼ŒSVMã®è¨“練データã«æ¯’を盛るãŸã‚ã®ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã‚’æ案 毒を盛るã¨ã¯ï¼Œã™ã§ã«å¦ç¿’ã•ã‚ŒãŸSVMã®ãƒ‘フォーマンスを劇的ã«ä¸‹ã’ã‚‹ãŸã‚ã®ãƒ‡ãƒ¼ã‚¿ã‚’生æˆã—ã¦ï¼ŒSVMã®æ–°ãŸãªæ•™å¸«ãƒ‡ãƒ¼ã‚¿ã¨ã—ã¦æ½œã‚Šè¾¼ã¾ã›ã‚‹ã“㨠手書ãæ–‡å—èªè˜ã§ã®å®Ÿé¨“ã§ã¯ï¼ŒãŸ
リリースã€éšœå®³æƒ…å ±ãªã©ã®ã‚µãƒ¼ãƒ“スã®ãŠçŸ¥ã‚‰ã›
最新ã®äººæ°—エントリーã®é…ä¿¡
j次ã®ãƒ–ックマーク
kå‰ã®ãƒ–ックマーク
lã‚ã¨ã§èªã‚€
eコメント一覧を開ã
oページを開ã
ディープラーニング ãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«ï¼ˆã‚‚ã—ãã¯ç ”究動å‘å ±å‘Šï¼‰
Deep learning with COTS HPC systems
Deep learning
Large Scale Machine Learning and Other Animals
Introduction To Machine Learning (Smola & Vishwanathan)
Amazon.com: Machine Learning: A Probabilistic Perspective (Adaptive Computation and Machine Learning series): Murphy, Kevin P.: Books
https://www.kurims.kyoto-u.ac.jp/~kyodo/kokyuroku/contents/pdf/1616-02.pdf
Amazon.co.jp: データã‹ã‚‰ã®çŸ¥è˜ç™ºè¦‹ (放é€å¤§å¦æ•™æ): 秋光淳生: 本
æ ªå¼ä¼šç¤¾Preferred Infrastructure
{{#tags}}- {{label}}
{{/tags}}