[B! LSH] yassã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

yass id:yass

LSHã«é–¢ã™ã‚‹yassã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (24)

${{author_name}}$

{{{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}}

Super-Bit LSH
ä»Šæ—¥ã¯ã€ã™ã“ã—å‰ã«èªã‚“ã Super-Bit LSH (SB-LSH)ã¨ã„ã†æ‰‹æ³•ã‚’ç°¡å˜ã«ç´¹ä»‹ã—ã¾ã™ã€‚ ã“ã‚Œã¯è§’é¡žä¼¼åº¦1ã«å¯¾ã™ã‚‹ãƒ©ãƒ³ãƒ€ãƒ å°„å½±LSH2ã®æ”¹è‰¯ç‰ˆã§ã™ã€‚ Jianqiu Ji, Jianmin Li, Shuicheng Yany, Bo Zhang, Qi Tianz. Super-Bit Locality-Sensitive Hashing. NIPS 2012. ãƒ©ãƒ³ãƒ€ãƒ å°„å½±LSHã®è§£èª¬ã¯æµ·é‡Žã•ã‚“ã®ã‚¹ãƒ©ã‚¤ãƒ‰(29-30ãƒšãƒ¼ã‚¸ç›®)ãŒã‚ã‹ã‚Šã‚„ã™ã„ã§ã™ã€‚ ãƒ©ãƒ³ãƒ€ãƒ å°„å½±LSHã‚’ä½¿ã†ã¨ã€è§’é¡žä¼¼åº¦ã®ä¸åæŽ¨å®šé‡ãŒå¾—ã‚‰ã‚Œã¾ã™ã€‚ Super-Bit LSHã§ã¯ã“ã‚Œã‚’Nãƒ“ãƒƒãƒˆã”ã¨ã«ã‚°ãƒ«ãƒ¼ãƒ—åˆ†ã‘ã—ã¦ã€å„ã‚°ãƒ«ãƒ¼ãƒ—å†…ã®å°„å½±ãƒ™ã‚¯ãƒˆãƒ«ã‚’ç›´äº¤ã•ã›ã‚‹ã¨ã„ã†ã‚‚ã®ã§ã™ã€‚ ç›´äº¤åŒ–ã¯ãŸã¨ãˆã°Gram-Schmidtã®ç›´äº¤åŒ–æ³•ã‚’ä½¿ã„ã¾ã™ã€‚ ã“ã†ã™ã‚‹ã¨ã€æŽ¨å®šé‡ã®åˆ†æ•£ãŒãŸã ã®ãƒ©ãƒ³ãƒ€ãƒ å°„å½±ã‚ˆã‚Šã‚‚å°ã•ããªã‚‹ã¨ã„ã†ã“ã¨ã§ã™ã€‚ è«–æ–‡
yass 2014/09/15
" ãƒ©ãƒ³ãƒ€ãƒ å°„å½±LSHã¯LSHã®ä¸ã§ã‚‚ã‹ãªã‚Šå¤å…¸çš„ãªã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã§ã™ãŒã€ãƒ‡ãƒ¼ã‚¿éžä¾å˜ã¨ã„ã†è¨å®šã§ãã®ç²¾åº¦ãŒæ”¹è‰¯ã•ã‚Œã‚‹ã®ã¯ãŠãã‚‰ãåˆã‚ã¦ã ã¨æ€ã„ã¾ã™ï¼ˆè«–æ–‡ä¸ã§ã‚‚ãã†æ›¸ã‹ã‚Œã¦ã„ã¾ã™ï¼‰"

lsh
ãƒªãƒ³ã‚¯
å¤§è¦æ¨¡ãƒ‡ãƒ¼ã‚¿ã®ãƒžã‚¤ãƒ‹ãƒ³ã‚° - å…±ç«‹å‡ºç‰ˆ
ã‚¦ã‚§ãƒ–ã‚„ã‚¤ãƒ³ã‚¿ãƒ¼ãƒãƒƒãƒˆã‚³ãƒžãƒ¼ã‚¹ã®éš†ç››ã«ã‚ˆã£ã¦ã€ã¨ã¦ã¤ã‚‚ãªãå·¨å¤§ãªãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆãŒå‡ºç¾ã—ã€ãã—ã¦ã€ãƒ‡ãƒ¼ã‚¿ãƒžã‚¤ãƒ‹ãƒ³ã‚°ã«ã‚ˆã£ã¦ãã‚Œã‚‰ã‹ã‚‰æƒ…å ±ãŒæŠ½å‡ºã•ã‚Œã¦ã„ã‚‹ã€‚ å·¨å¤§ãªãƒ‡ãƒ¼ã‚¿ã«å¯¾ã™ã‚‹ãƒžã‚¤ãƒ‹ãƒ³ã‚°ã«ãŠã„ã¦ã¯ã€å¾“æ¥ã‹ã‚‰çŸ¥ã‚‰ã‚Œã¦ã„ã‚‹ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ãŒãã®ã¾ã¾ã§ã¯æ©Ÿèƒ½ã—ãªã„ã“ã¨ãŒå¤šã„ã€‚å·¨å¤§ãªãƒ‡ãƒ¼ã‚¿ã‚’æ‰±ã†ä¸Šã§ã“ã‚Œã¾ã§ã¨ã¯é•ã£ãŸå‡¦ç†æˆ¦ç•¥ã‚„æŠ€æ³•ãŒå¿…è¦ã¨ãªã‚‹ã€‚ æœ¬æ›¸ã¯ãƒ‡ãƒ¼ã‚¿ãƒžã‚¤ãƒ‹ãƒ³ã‚°ã«ãŠã„ã¦é‡è¦ãªå•é¡Œã®è§£æ±ºã«ä½¿ã‚ã‚Œã¦ãã¦ãŠã‚Šã€ã•ã‚‰ã«å·¨å¤§ãªãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã«å¯¾ã—ã¦ã‚‚ä½¿ç”¨ã§ãã‚‹å®Ÿç”¨çš„ãªã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã‚’ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚„ã‚¦ã‚§ãƒ–æŠ€è¡“ã®åˆ†é‡Žã§è‘—åãªåŽŸè‘—è€…ãŒè§£èª¬ã™ã‚‹ã€‚ ã¾ãšã€ãƒ‡ãƒ¼ã‚¿ãƒžã‚¤ãƒ‹ãƒ³ã‚°ã®æœ¬è³ªã‚„ã€ãƒ‡ãƒ¼ã‚¿ãƒžã‚¤ãƒ‹ãƒ³ã‚°ãŒã©ã®ã‚ˆã†ã«æ‰±ã‚ã‚Œã¦ã„ã‚‹ã‹ã‚’æ¦‚è¦³ã™ã‚‹ã€‚æ¬¡ã«ã€ä»Šå¾Œéžå¸¸ã«å¤§é‡ã®ãƒ‡ãƒ¼ã‚¿ã®è§£æžã‚’è¡Œã†éš›ã«ã€ã‚¯ãƒ©ã‚¦ãƒ‰ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã¨ã¨ã‚‚ã«é‡è¦è¦–ã•ã‚Œã‚‹ã¨è€ƒãˆã‚‰ã‚Œã€ã“ã®å¾Œã®ç« ã®è°è«–ã«å¿…é ˆã®ã‚‚ã®ã¨ãªã‚‹ãƒžãƒƒãƒ—ãƒ¬ãƒ‡ãƒ¥ãƒ¼ã‚¹ã‚’è§£èª¬ã™ã‚‹ã€‚ãã®å¾Œã€é¡žä¼¼ã™ã‚‹ã‚¢ã‚¤ãƒ†ãƒ
yass 2014/07/01
book

LSH

minhash

MapReduce

PageRank

machinelearning

recommend
ãƒªãƒ³ã‚¯
Multi-Index Locality Sensitive Hashing for Fun and Profit - Engineering Blog
yass 2014/06/01
minhash

lsh

similarity
ãƒªãƒ³ã‚¯
GoogleNewsã®ãƒ¬ã‚³ãƒ¡ãƒ³ãƒ‰ã®ä¸èº« - UMEko Branding
å…ˆæ—¥ã€å…¨ä½“ã‚¼ãƒŸã§ç™ºè¡¨ã—ãŸã¨ãã®å†…å®¹ã§ã™ãŒã€ã“ã“ã«ã¾ã¨ã‚ã¨ãã¾ã™ã€‚ã€‚GoogleNewsã®ãƒ¬ã‚³ãƒ¡ãƒ³ãƒ‰ã®ä¸èº«ã‚’è¿½ã£ãŸè«–æ–‡ã®è¦ç´„ã§ã™ã€‚å°‘ã—å‰ã®å…¨ä½“ã‚¼ãƒŸã§ç”¨ã„ãŸè³‡æ–™ã§ã™ã€‚ã‚½ãƒ¼ã‚¹ï¼šAbhinandan Das,Mayur Datar,Ashutosh Garg,Shyam Rajaram,"Google News Personalization: Scala ble OnlineCollaborative Filtering",WWW2007ä¸å‹‰å¼·ãªå€‹æ‰€ãŒå¤šã€…ã‚ã‚Šã¾ã™ã®ã§ã€èª¤ã£ã¦ã„ã‚‹ç®‡æ‰€ç‰ã‚ã‚Šã¾ã—ãŸã‚‰ã€æ˜¯éžã”æŒ‡æ‘˜ãã ã•ã„ã€‚ å€‹äººçš„ã«ã¯ã€æœ€è¿‘ã®ãƒ¢ãƒ‡ãƒ«ãƒ™ãƒ¼ã‚¹ã®æ‰‹æ³•ã®å‹‰å¼·ãƒ»ãŠã•ã‚‰ã„ã¨ã„ã†æ„å‘³ã§ç”¨ã„ã¦ã„ã‚‹ã®ã§ã€GoogleNewsç‹¬è‡ªã®æ‹¡å¼µãªã‚Šå®Ÿè£…ã®éƒ¨åˆ†ã®å†…å®¹ãŒçœã‹ã‚Œã¦ã„ã‚‹å ´åˆãŒã‚ã‚Šã¾ã™ã€‚ã¾ãŸã€ãƒ‡ãƒ¼ã‚¿æ§‹é€ ã‚„MapReduceã‚’ç”¨ã„ãŸè¨ˆç®—ã®ä»•çµ„ã¿ã®éƒ¨åˆ†ã¯ã€ã“ã“ã§ã¯çœç•¥ã—ã¦ã„ã¾ã™ã€‚ã€‚ä¸€å¿œã€ å…¨ä½“åƒã€€ãƒ»LSH(Lo
yass 2013/05/11
recommend

google news

lsh

minhash

plsi
ãƒªãƒ³ã‚¯
Locality Sensitive Hashã«ã‚ˆã‚‹é¡žä¼¼ãƒ™ã‚¯ãƒˆãƒ«æ¤œç´¢ã‚’è©¦ã™ - Negative/Positive Thinking
ã¯ã˜ã‚ã« é¡žä¼¼æ€§ãŒé«˜ã„ãƒ™ã‚¯ãƒˆãƒ«ã®ãƒãƒƒã‚·ãƒ¥å€¤ãŒè¿‘ã„å€¤ã«ãªã‚‹ã‚ˆã†ãªãƒãƒƒã‚·ãƒ¥é–¢æ•°ã‚’ä½¿ã£ã¦ã€ é¡žä¼¼ã™ã‚‹ã‚‚ã®ã‚’é«˜é€Ÿã«æ¤œç´¢ã™ã‚‹ã“ã¨ãŒã§ãã‚‹ã®ã§ã€ãã‚Œã‚’è©¦ã—ã¦ã¿ãŸã€‚ Locality Sensitive Hash é¡žä¼¼ã™ã‚‹ãƒ‡ãƒ¼ã‚¿ãŒé«˜ç¢ºçŽ‡ã§è¿‘ã„å€¤ã«ãªã‚‹(Locality-Sensitive)ãƒãƒƒã‚·ãƒ¥é–¢æ•°ã®ã“ã¨ é«˜æ¬¡å…ƒãƒ‡ãƒ¼ã‚¿ã®æ¬¡å…ƒåœ§ç¸®ã‚’è¡Œãˆã‚‹ (P1,P2,r,cr)-sensitiveãªHashæ—ã¨ã¯ã€ 2ã¤ã®ç‰¹å¾´ãƒ™ã‚¯ãƒˆãƒ«p,qã«ã¤ã„ã¦(P1>P2) ||p-q||P1 ||p-q||>crãªã‚‰Pr[h(p)=h(q)] ã‚’æº€ãŸã™ãƒãƒƒã‚·ãƒ¥é–¢æ•°h:R^d->U ã‚³ã‚µã‚¤ãƒ³é¡žä¼¼åº¦ã«å¯¾ã™ã‚‹LSH 2ã¤ã®kæ¬¡å…ƒãƒ™ã‚¯ãƒˆãƒ«u,vã«ã¤ã„ã¦ ã‚³ã‚µã‚¤ãƒ³é¡žä¼¼åº¦: u*v / sqrt(|u|*|v|) då€‹ã®kæ¬¡å…ƒã®ãƒ©ãƒ³ãƒ€ãƒ ãƒ™ã‚¯ãƒˆãƒ«r_iã‚’è€ƒãˆã€ãƒãƒƒã‚·ãƒ¥é–¢æ•°h_i(u)ã‚’ h_i(u) = 1 (r*u >=0) h_i(u)
yass 2013/05/01
LSH

similarity
ãƒªãƒ³ã‚¯
Google Code Archive - Long-term storage for Google Code Project Hosting.
Code Archive Skip to content Google About Google Privacy Terms
yass 2013/02/11
LSH

java

simhash

random projection
ãƒªãƒ³ã‚¯
Stanford CS345A: Data Mining
Handouts: 1/5: Introduction Introduction [slides] [reading] 1/7: MapReduce MapReduce [slides] [reading] 1/12: Recommendation System Recommendation System [slides] [reading] 1/14: Near Neighbor Search in High Dimensional Data Near Neighbor Search in High Dimensional Data [slides] [reading] 1/19: Locality Sensitive Hashing (LSH) Locality Sensitive Hashing (LSH) [slides] [reading] 1/21: Structure of
yass 2013/02/11
2010

data mining

LSH

recommend

PageRank

stanford
ãƒªãƒ³ã‚¯
FAST LARGE-SCALE APPROXIMATE GRAPH CONSTRUCTION FOR NLP
yass 2013/02/11
LSH

Count-Min Sketch

similarity

sort
ãƒªãƒ³ã‚¯
è¤‡åˆã‚½ãƒ¼ãƒˆæ³•ã«ã‚ˆã‚‹é«˜é€Ÿãªå…¨ãƒšã‚¢ é¡žä¼¼åº¦æ¤œç´¢
yass 2013/02/11
sketch sort

LSH

similarity
ãƒªãƒ³ã‚¯
ä¹±æŠžã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã¨è‡ªç„¶è¨€èªžå‡¦ç†: Locality Sensitive Hashing ã‚’ç”¨ã„ãŸ é«˜é€Ÿãªåè©žã‚¯ãƒ©ã‚¹ã‚¿ãƒªãƒ³ã‚°
Company Logo ä¹±æŠžã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã¨è‡ªç„¶è¨€èªžå‡¦ç†: Locality Sensitive Hashing ã‚’ç”¨ã„ãŸ é«˜é€Ÿãªåè©žã‚¯ãƒ©ã‚¹ã‚¿ãƒªãƒ³ã‚° Randomized Algorithms and NLP: Using Locality Sensitive Hash Functions for High Speed Noun Clustering D. Ravichandran, P. Pantel, and E. Hovy (ACL 2005) èªã¿â¼¿: å²¡å´Žç›´è¦³ Company Logo ã“ã®è«–æ–‡ã«ã¤ã„ã¦ â¼¿æ³•ã¨ãã®èƒŒæ™¯ç†è«–ãŒã™ã”ãèˆˆå‘³æ·±ã„ â€“ ä»–ã«ã‚‚ã„ã‚ã„ã‚ä½¿ãˆãã† ä¹±æŠžã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã®é›°å›²æ°—ãŒå‘³ã‚ãˆã‚‹ â€“ ãƒ©ãƒ³ãƒ€ãƒ ãŒç¹”ã‚Šãªã™ä¸æ€è°ãªä¸–ç•Œ ã™ãã‚ŒãŸå®Ÿè£…â½…æ³•ã‚’â½°ã—ã¦ã„ã‚‹ â€“ èª°ã§ã‚‚ç°¡å˜ã«å®Ÿè£…ã§ãã‚‹ 2005å¹´ã®ç ”ç©¶ã ãŒï¼Œã„ã¾ã ã«â¾Šè¤ªã›ã¦ã¯ã„ãªã„ â€“ ç”»åƒé¡žä¼¼æ¤œç´¢ã§Lo
yass 2013/02/11
LSH

nlp

similarity
ãƒªãƒ³ã‚¯
Hadoop Streaming ã§ Locality-Sensitive Hashing ã‚’å®Ÿè£…ã—ã¦ã¿ã‚‹ | Atsushi TATSUMA Web Page
ã¯ã˜ã‚ã« ç ”ç©¶å®¤ã§ã¯ã€å¤§è¦æ¨¡ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’å¯¾è±¡ã¨ã—ãŸæ¤œç´¢ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã«ã¤ã„ã¦ç ”ç©¶ã—ã¦ã„ã¾ã™ã€‚ åˆ†æ•£å‡¦ç†ã«ã‚ˆã‚‹ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ä½œæˆã‚’è€ƒãˆã¦ã„ã¦ã€ä»Šå›ž Hadoop Streaming ã§ã©ã“ã¾ã§ã§ãã‚‹ã‹ã‚’è©¦ã™ã¹ãã€ Locality-Sensitive Hashing (LSH) ã‚’å®Ÿè£…ã—ã¦ã¿ã¾ã—ãŸã€‚ å®Ÿè£…ã—ãŸã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã«ã¤ã„ã¦ LSH ã«ã¯ã„ãã¤ã‹ã®ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã®ãƒãƒªã‚¨ãƒ¼ã‚·ãƒ§ãƒ³ãŒã‚ã‚Šã¾ã™ã€‚ LSH ã®è©³ç´°ã¯ã€ãƒ–ãƒã‚°ãªã‚Šè«–æ–‡ãªã‚Šæœ¬ãªã‚Šã‚ã‚Šã¾ã™ã®ã§ã€ã“ã“ã§ã¯çœç•¥ã—ã¾ã™ãŒã€ é¡žä¼¼ã—ãŸãƒ‡ãƒ¼ã‚¿ã«åŒã˜ãƒãƒƒã‚·ãƒ¥å€¤ã‚’ä¸Žãˆã‚‹ã“ã¨ã§ã€æ¤œç´¢ã‚’é«˜é€ŸåŒ–ã—ã‚ˆã†ã¨ã„ã†ã‚¢ã‚¤ãƒ‡ã‚£ã‚¢ã§ã™ã€‚ ã“ã®ãƒãƒƒã‚·ãƒ¥å€¤ã«ã¯ã€0101 ã¨ã‹ã®çŸã„ãƒã‚¤ãƒŠãƒªç¬¦å·ãŒå¥½ã¾ã‚Œã¾ã™ã€‚ ä»Šå›žã¯ã€ã„ãã¤ã‹ã‚ã‚‹ LSH ã®ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã®ã†ã¡ã€ Charikar,M., Similarity estimation techniques from ro
yass 2013/02/11
LSH

hadoop

python
ãƒªãƒ³ã‚¯
GitHub - bbc/ruby-lsh: Locality Sensitive Hashing in Ruby
You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session. You switched accounts on another tab or window. Reload to refresh your session. Dismiss alert
yass 2013/02/11
LSH

ruby

redis
ãƒªãƒ³ã‚¯
3æ—¥ã§ä½œã‚‹é«˜é€Ÿç‰¹å®šç‰©ä½“èªè˜ã‚·ã‚¹ãƒ†ãƒ (7) æœ€è¿‘å‚æŽ¢ç´¢ã®é«˜é€ŸåŒ– - äººå·¥çŸ¥èƒ½ã«é–¢ã™ã‚‹æ–å‰µéŒ²
3æ—¥ã§ä½œã‚‹é«˜é€Ÿç‰¹å®šç‰©ä½“èªè˜ã‚·ã‚¹ãƒ†ãƒ (6) ç·šå½¢æŽ¢ç´¢ã‚’ç”¨ã„ãŸç‰¹å®šç‰©ä½“èªè˜ï¼ˆ2009/11/22ï¼‰ã®ã¤ã¥ãã§ã™ã€‚ä»Šå›žãŒã“ã®ã‚·ãƒªãƒ¼ã‚ºã®æœ€çµ‚å›žã§ã™ã€‚ å‰å›žã®ç·šå½¢æŽ¢ç´¢ã¯é…ã™ãŽã‚‹ã®ã§æœ€è¿‘å‚æŽ¢ç´¢ã‚’é«˜é€ŸåŒ–ã—ã¾ã™ã€‚ã“ã‚Œã§è¡¨é¡Œã®é«˜é€Ÿç‰¹å®šç‰©ä½“èªè˜ã‚·ã‚¹ãƒ†ãƒ ãŒã§ãã‚ãŒã‚Šã¾ã™ã€‚é«˜é€ŸåŒ–ã«ã¯ã„ãã¤ã‹ã®æ–¹æ³•ãŒã‚ã‚Šã¾ã™ãŒã€ç‰©ä½“ãƒ¢ãƒ‡ãƒ«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’ãªã‚“ã‚‰ã‹ã®ãƒ‡ãƒ¼ã‚¿æ§‹é€ ã«ã‚ã‚‰ã‹ã˜ã‚æ ¼ç´ã—ã¦ãŠãã¨ã„ã†ã®ãŒãƒã‚¤ãƒ³ãƒˆã§ã™ã€‚ä»Šå›žã¯ã€è³‡æ–™ã§ã‚‚è¿°ã¹ã‚‰ã‚Œã¦ã„ã‚‹kd-treeã¨Locality Sensitive Hashing (LSH)ã¨ã„ã†æ‰‹æ³•ã‚’è©¦ã—ã¦ã¿ã¾ã™ã€‚kd-treeã¯æœ¨æ§‹é€ ã€LSHã¯ãƒãƒƒã‚·ãƒ¥ã§ãƒ‡ãƒ¼ã‚¿ã‚’æ§‹é€ åŒ–ï¼ˆã‚¤ãƒ³ãƒ‡ã‚ã‚·ãƒ³ã‚°ï¼‰ã—ã¾ã™ã€‚kd-treeã¯ã€åŽ³å¯†ãªæœ€è¿‘å‚ã‚’æ±‚ã‚ã¾ã™ãŒã€LSHã¯è¿‘ä¼¼æœ€è¿‘å‚æ¤œç´¢ã¨å‘¼ã°ã‚Œã€åŽ³å¯†ãªæœ€è¿‘å‚ã¯æ±‚ã‚ã‚‰ã‚Œãªã„ä»£ã‚ã‚Šã«è¨ˆç®—ã‚’å¤§å¹…ã«é«˜é€ŸåŒ–ã§ãã¾ã™ã€‚ è³‡æ–™ã§ã¯ã€ANN (Approximate Nearest
yass 2013/02/11
LSH

kd-tree

opencv
ãƒªãƒ³ã‚¯
b-Bit Minwise Hashing
yass 2013/02/10
similarity

minhash

LSH

hash
ãƒªãƒ³ã‚¯
Learning to Hash! æœ€æ–° Locality sensitive hashing äº‹æƒ… - æ¦è”µé‡Žæ—¥è¨˜
é«˜é€Ÿã«é¡žä¼¼åº¦è¨ˆç®—ã‚’ã—ãŸã„å ´åˆã€å…¸åž‹çš„ã«ä½¿ã‚ã‚Œã‚‹ã®ã¯ Locality sensitive hashing (LSH)ã¨ã„ã†æŠ€è¡“ã§ã‚ã‚Šã€å…ƒã€…è·é›¢ãŒè¿‘ã„ã‚¤ãƒ³ã‚¹ã‚¿ãƒ³ã‚¹åŒå£«ã¯ãƒãƒƒã‚·ãƒ¥å€¤ãŒè¿‘ããªã‚‹ã‚ˆã†ã«ãƒãƒƒã‚·ãƒ¥é–¢æ•°ã‚’ä½œã‚‹ã“ã¨ã§é«˜é€Ÿã«é¡žä¼¼åº¦ã‚’è¨ˆç®—ã—ãŸã‚Šã§ãã‚‹ã¨ã„ã†ãŠè©±ãªã®ã ãŒã€æœ€è¿‘ Semantic hashing ã‚„ Spectral hashingã€ã¾ãŸ Kernelized LSH ã¨ã„ã†æ‰‹æ³•ãŒç™»å ´ã—ã¦ç››ã‚Šä¸ŠãŒã‚Šã¤ã¤ã‚ã‚‹ã¨ã“ã‚ã€åŒã˜ã‚°ãƒ«ãƒ¼ãƒ—ã®äººãŒã‚‚ã£ã¨ã„ã„ã®ã‚’å‡ºã—ã¾ã—ãŸã€ã¨ã„ã†ã“ã¨ã‚‰ã—ã„ã€‚ã¡ãªã¿ã«æƒ…å ±æŽ¨è–¦ã¨ã‹ç”»åƒæ¤œç´¢ã¨ã‹å¤§è¦æ¨¡ã‚¯ãƒ©ã‚¹ã‚¿ãƒªãƒ³ã‚°ã¨ã‹ã€ã„ã‚ã„ã‚ãªåˆ†é‡Žã§é«˜é€Ÿãªé¡žä¼¼åº¦è¨ˆç®—ã®å¿œç”¨ä¾‹ãŒã‚ã‚‹ã€‚ ãã†ã„ã†ã‚ã‘ã§ã€ä»Šæ—¥ã¯ manab-ki ãã‚“ãŒ Brian Kulis and Trevor Darrell. "Learning to Hash with Binary Reconstructive
yass 2013/02/03
hash

LSH

Spectral hashing
ãƒªãƒ³ã‚¯
Locality Sensitive Hashing
ã€è«–æ–‡ã€‘M. Datar, N. Immorlica, P. Indyk, and V.S. Mirrokni, Â“Locality-Sensitive Hashing Scheme Based on p-Stable DistributionsÂ”, Proceedings of the twentieth annual symposium on Computational geometry, 2004. Overview Locality Sensitive Hashing(ä»¥ä¸‹LSHï¼Œå±€æ‰€æ€§é‹æ•åž‹ãƒãƒƒã‚·ãƒ¥)ã¯é«˜æ¬¡å…ƒãƒ‡ãƒ¼ã‚¿ã‚’æ¬¡å…ƒåœ§ç¸®ã™ã‚‹ãŸã‚ã®æ‰‹æ³•ã§ï¼Œåœ§ç¸®ã—ãŸãƒ‡ãƒ¼ã‚¿ã«ã‚ˆã‚‹ãƒãƒƒã‚·ãƒ¥ãƒ†ãƒ¼ãƒ–ãƒ«ã‚’ä½œæˆã—ã¦ï¼ŒåŠ¹çŽ‡çš„ã«è¿‘ä¼¼çš„æœ€è¿‘å‚æŽ¢ç´¢ã‚’è¡Œã†ãŸã‚ã®æ‰‹æ³•ã§ã™ï¼ŽåŸºæœ¬çš„ãªã‚¢ã‚¤ãƒ‡ã‚£ã‚¢ã¨ã—ã¦ã¯é«˜æ¬¡å…ƒç©ºé–“å†…ã§ã®è·é›¢ãŒè¿‘ã„2ã¤ã®ãƒ™ã‚¯ãƒˆãƒ«ãŒï¼Œãƒãƒƒã‚·ãƒ¥ãƒ†ãƒ¼ãƒ–ãƒ«ã«ãŠã„ã¦é«˜ç¢ºçŽ‡ã§åŒã˜ãƒã‚±ãƒƒãƒˆã«å…¥ã‚‹ã‚ˆã†ã«ç¢ºçŽ‡çš„ãªå‡¦ç†ã‚’è¡Œã†ã¨
yass 2013/02/03
hash

LSH

java

kd-tree
ãƒªãƒ³ã‚¯
ãƒ¬ã‚³ãƒ¡ãƒ³ãƒ‰, LSH, Spectral Hashing - DO++
WEB+DB press vol.49ã«ãƒ¬ã‚³ãƒ¡ãƒ³ãƒ‰ç‰¹é›†ã®è¨˜äº‹ã‚’tkngã•ã‚“ã¨æ›¸ãã¾ã—ãŸã€‚ å†…å®¹ã¯æœ€åˆã¯ã€å”èª¿ãƒ•ã‚£ãƒ«ã‚¿ãƒªãƒ³ã‚°ã‚„ã‚³ãƒ³ãƒ†ãƒ³ãƒ„ãƒžãƒƒãƒã®ç°¡å˜ãªè©±ã‹ã‚‰ã€ç‰¹å¾´é‡ã‚’ã©ã®ã‚ˆã†ã«è¡¨ã™ã‹ã€å¤§è¦æ¨¡ãƒ‡ãƒ¼ã‚¿ã‚’ã©ã®ã‚ˆã†ã«å‡¦ç†ã™ã‚‹ã‹ã«ã„ãã€ç‰¹ç•°å€¤åˆ†è§£ãªã©ã®ä½Žãƒ©ãƒ³ã‚¯è¡Œåˆ—åˆ†è§£ã«ã‚ˆã‚‹ãƒ¬ã‚³ãƒ¡ãƒ³ãƒ‰ã‚„Restricted Boltzmann Machineã¨ã„ã£ãŸæœ€è¿‘ã®netflix prizeã®ä¸Šä½ã®æ‰‹æ³•ãªã©ã€ã‹ãªã‚Šçªã£è¾¼ã‚“ã è°è«–ã‚‚ã—ã¦ã¾ã™ã€‚ å€‹äººçš„ã«ã¯ä¸‰ç« ã§Locality Sensitive Hash(LSH)ã«ã¤ã„ã¦æ‰±ã£ã¦ã„ã‚‹ã‚ãŸã‚ŠãŒãŠå‹§ã‚ã§ã™ã€‚ ãƒ¬ã‚³ãƒ¡ãƒ³ãƒ‰ã®å†…éƒ¨ã®å•é¡Œã‚’æ¥µè¨€ã™ã‚‹ã¨ã€ãƒ‡ãƒ¼ã‚¿ã¨ã„ã†ã®ã¯ç–Žãªé«˜æ¬¡å…ƒã®æ•°å€¤ãƒ™ã‚¯ãƒˆãƒ«ï¼ˆæ•°ç™¾ä¸‡æ¬¡å…ƒã¨ã‹ï¼‰ã§è¡¨ã•ã‚Œã€ã‚¯ã‚¨ãƒªã§ãƒ™ã‚¯ãƒˆãƒ«ãŒä¸Žãˆã‚‰ã‚ŒãŸæ™‚ã€ã“ã‚Œã¨ä¼¼ãŸã‚ˆã†ãªãƒ™ã‚¯ãƒˆãƒ«ã‚’æŽ¢ã—ã¦ã“ã„ã¨ã„ã†å•é¡Œã«ãªã‚Šã¾ã™ã€‚â€ä¼¼ãŸã‚ˆã†ãªâ€ã‚’æ•°å¦çš„ã«ã„ãˆã°ã€ã‚¯ã‚¨ãƒªã®ãƒ™ã‚¯ãƒˆãƒ«ã¨ã®å†…ç©ï¼ˆå„ãƒ™ã‚¯ãƒˆãƒ«ã¯é•·
yass 2013/02/03
hash

LSH

Spectral hashing

recommend

book
ãƒªãƒ³ã‚¯
PFI Seminar 2012/03/15 ã‚«ãƒ¼ãƒãƒ«ã¨ãƒãƒƒã‚·ãƒ¥ã®æ©Ÿæ¢°å¦ç¿’
2. è‡ªå·±ç´¹ä»‹ å¾—å±… èª ä¹Ÿ | Tokui, Seiya æ±äº¬å¤§å¦ ç†å¦éƒ¨ æ•°å¦ç§‘ @beam2d åŒ æƒ…å ±ç†å·¥ æ•°ç†æƒ…å ±å¦ M2 è¿‘å‚æŽ¢ç´¢ã®ãŸã‚ã®ãƒãƒƒã‚·ãƒ¥ã®å¦ç¿’ PFIã‚¢ãƒ«ãƒã‚¤ãƒˆï¼ˆä¸€å¹´åŠãã‚‰ã„ï¼‰ æ¥æœˆã‹ã‚‰æ£ç¤¾å“¡ 3. ä»Šæ—¥ã®ãŠã¯ãªã— ãƒãƒƒã‚·ãƒ¥ â€¢ è¿‘å‚æŽ¢ç´¢ï¼šã‚¯ã‚¨ãƒªã«é¡žä¼¼ã—ãŸãƒ‡ãƒ¼ã‚¿ã‚’ åŠ¹çŽ‡è‰¯ãå–ã‚Šå‡ºã™ â€¢ æ©Ÿæ¢°å¦ç¿’ï¼šåˆ†é¡žã‚„å›žå¸°ã‚’åŠ¹çŽ‡è‰¯ãå¦ç¿’ã™ã‚‹ ã‚ã‚‹ã¨è‰¯ã„çŸ¥è˜ â€¢ ç·šå½¢å¦ç¿’å™¨ã‚’ä¸€ã¤ã§ã‚‚çŸ¥ã£ã¦ã„ã‚‹ã¨ ã‚ã‹ã‚Šã‚„ã™ã„ã¨æ€ã„ã¾ã™ 4. å†…å®¹ â€¢ ã‚«ãƒ¼ãƒãƒ«ã®ç°¡å˜ãªèª¬æ˜Žï¼ˆå¾©ç¿’ï¼‰ â€¢ ãƒãƒƒã‚·ãƒ¥é–¢æ•°ã®å°Žå…¥ â€¢ è¿‘å‚æŽ¢ç´¢ã«ä½¿ãˆã‚‹ãƒãƒƒã‚·ãƒ¥ â€“ ãƒ‡ãƒ¼ã‚¿ã®ãƒãƒƒã‚·ãƒ¥ â€“ ã„ã‚ã„ã‚ãªæ‰‹æ³•ã‚’ç´¹ä»‹ (LSHç³», ã‚«ãƒ¼ãƒãƒ«, ãƒãƒƒã‚·ãƒ¥ã®å¦ç¿’) â€¢ æ©Ÿæ¢°å¦ç¿’ã«ä½¿ãˆã‚‹ãƒãƒƒã‚·ãƒ¥ â€“ ç´ æ€§ã®ãƒãƒƒã‚·ãƒ¥ â€“ ãƒãƒƒã‚·ãƒ¥ã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ã®å¦ç¿’
yass 2013/02/03
hash

minhash

Spectral hashing

LSH

hashing trick
ãƒªãƒ³ã‚¯
æœ€è¿‘ã®ãƒã‚¤ãƒŠãƒªãƒãƒƒã‚·ãƒ³ã‚°ã‚’ã„ãã¤ã‹Javaã§å®Ÿè£…ã—ã¦ã¿ãŸ - rubyu's blog
åŽ»å¹´ã®çµ‚ã‚ã‚Šã‹ã‚‰ã€ãƒã‚¤ãƒŠãƒªãƒãƒƒã‚·ãƒ³ã‚°ã‚’ä½¿ã£ãŸè¿‘ä¼¼è¿‘å‚æ¤œç´¢ã‚’ã„ã‚ã„ã‚èª¿ã¹ã¦ã„ãŸã®ã§ã™ãŒã€ã¼ã¡ã¼ã¡ä¸€æ®µè½ã—ãŸã®ã§ã€ã²ã¨é€šã‚Šã¾ã¨ã‚ã¦ãŠãã¾ã™ã€‚ ãƒã‚¤ãƒŠãƒªãƒãƒƒã‚·ãƒ³ã‚°ã¨ã¯ã€‚ å€‹ã® æ¬¡å…ƒã®ç‚¹ã‹ã‚‰ãªã‚‹ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆ ã§ã€å…ƒç©ºé–“ã§ã®è¿‘å‚ç‚¹ã‚’ã€é¡žä¼¼ã—ãŸãƒã‚¤ãƒŠãƒªã‚³ãƒ¼ãƒ‰ã«é–¢é€£ã¥ã‘ã‚‹æŠ€è¡“ã€‚ è¦ã™ã‚‹ã«ã€å®Ÿæ•°ãƒ™ã‚¯ãƒˆãƒ«ã®æ¤œç´¢ã‚’ãƒžãƒˆãƒ¢ã«ã‚„ã‚‹ã«ã¯ã€æœ€è¿‘ã®ãƒ‡ãƒ¼ã‚¿ã¯è†¨å¤§ã™ãŽã‚‹ã®ã§ãŠæ‰‹ä¸Šã’ã€‚ãªã®ã§ã€å…ƒç©ºé–“ã§ã®è·é›¢ã‚’ãªã‚‹ã¹ãä¿ã£ãŸã¾ã¾ã€ãƒã‚¤ãƒŠãƒªã‚³ãƒ¼ãƒ‰ã«è½ã¨ã—ã¾ã—ã‚‡ã†ã€‚ ãã†ã™ã‚‹ã¨ã€ãƒã‚¤ãƒŠãƒªä¸€è‡´ã‹ã€1ãƒ“ãƒƒãƒˆé•ã†ã‹ã€2ãƒ“ãƒƒãƒˆé•ã†ã‹...ã¨ã€æœç´¢ã—ã¦ã„ãã«ã—ã¦ã‚‚ã€å…ƒç©ºé–“ã®ãƒ‡ãƒ¼ã‚¿ã§ã‚„ã‚‹ã‚ˆã‚Šé«˜é€Ÿã§ã€ã—ã‹ã‚‚ã‚¹ãƒˆãƒ¬ãƒ¼ã‚¸å®¹é‡ã‚’å‰Šæ¸›ã§ãã‚‹ã¨ã„ã†ã‚ã‘ã§ã™ã€‚ ãã® ãƒ“ãƒƒãƒˆã®ãƒã‚¤ãƒŠãƒªã‚³ãƒ¼ãƒ‰ ã‚’ä½œã‚‹ãŸã‚ã«ã€ å€‹ã®ãƒãƒƒã‚·ãƒ¥é–¢æ•°ãŒä½¿ã‚ã‚Œã‚‹ã€‚ ãƒãƒƒã‚·ãƒ¥é–¢æ•°ã¯ ã¨å®šç¾©ã•ã‚Œã‚‹ã€‚ã“ã“ã§ã€ ã¯ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã€‚ ã¯å°„å½±ãƒ™ã‚¯ãƒˆãƒ«ã€‚ ã¯é–¾å€¤ã€‚ ç·šå½¢å†™åƒãƒ™ãƒ¼ã‚¹ã®ãƒãƒƒã‚·ãƒ³ã‚°ã¯ã‚·ãƒ³ãƒ—ãƒ«
yass 2013/01/10
java

lsh

hash

Spectral hashing
ãƒªãƒ³ã‚¯
Similarity Join Algorithms: An Introduction
yass 2012/11/28
similarity

join

algorithm

toread

LSH

minhash
ãƒªãƒ³ã‚¯
1 2 æ¬¡ã®ãƒšãƒ¼ã‚¸