[B! ml] syou6162ã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

syou6162 id:syou6162

mlã«é–¢ã™ã‚‹syou6162ã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (34)

${{author_name}}$

{{{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}}

å‹¾é…ãƒ–ãƒ¼ã‚¹ãƒ†ã‚£ãƒ³ã‚°ã®åŸºç¤Žã¨æœ€æ–°ã®å‹•å‘ (MIRU2020 Tutorial)
Gradient Boostingã¯è¿‘å¹´Kaggleãªã©ã®ã‚³ãƒ³ãƒšãƒ†ã‚£ã‚·ãƒ§ãƒ³ã§æ³¨ç›®ã‚’é›†ã‚ã¦ã„ã‚‹åˆ†é¡žã‚„å›žå¸°å•é¡Œã«å¯¾ã™ã‚‹ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã®ä¸€ã¤ã§ã‚ã‚‹ã€‚XGBoost, LightGBM, CatBoostãªã©ãŒæœ‰åã§ã¯ã‚ã‚‹ãŒã€ãã‚Œã‚‰ã‚’åœŸå°ã«ã—ã¦è¿‘å¹´ã¯DeepGBMã‚„NGBoostã¨ã„ã£ãŸæ–°è¦ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã®ç™»å ´ã€æ±ŽåŒ–ç†è«–è§£æžã®é€²å±•ã€ãƒ¢ãƒ‡ãƒ«è§£é‡ˆæ€§ã®ä»˜ä¸Žæ–¹æ³•ã®å¤šæ§˜åŒ–ãªã©ã€ç†è«–ã‹ã‚‰å¿œç”¨ã¾ã§å¤šå²ã«ã‚ãŸã‚‹ç ”ç©¶ãŒè¡Œã‚ã‚Œã¦ã„ã‚‹ã€‚æœ¬ãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«ã§ã¯ã€Gradient Boostingã«é–¢ã™ã‚‹è¿‘å¹´ã®ç ”ç©¶å‹•å‘ã‚„ãƒ†ã‚¯ãƒ‹ãƒƒã‚¯ã‚’ã€ãã‚Œã‚‰ã®ç¤¾ä¼šå®Ÿè£…ã¾ã§ã‚’è¦‹æ®ãˆãªãŒã‚‰ç´¹ä»‹ã—ã¦ã„ãã€‚ Read less
syou6162 2020/08/02
ml
ãƒªãƒ³ã‚¯
ç¤¾å†…ã§å›½éš›ä¼šè°è«–æ–‡èªã¿ä¼šã‚’é–‹å‚¬ã—ã¾ã—ãŸ - Hatena Developer Blog
ã“ã‚“ã«ã¡ã¯ã€ã‚¦ã‚§ãƒ–ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã‚¨ãƒ³ã‚¸ãƒ‹ã‚¢ã®id:syou6162ã§ã™ã€‚ç§‹ã¾ã£ã•ã‹ã‚Šã€èªæ›¸ã®ç§‹/è«–æ–‡èªã¿ä¼šã®ç§‹ã§ã™ãã€‚ã¯ã¦ãªç¤¾å†…ã§å›½éš›ä¼šè°è«–æ–‡èªã¿ä¼šã‚’é–‹å‚¬ã—ãŸã®ã§ã€ä»Šå›žã¯èªã¿ä¼šã§å–ã‚Šä¸Šã’ãŸè«–æ–‡ã«ã¤ã„ã¦ç°¡å˜ã«ç´¹ä»‹ã—ã¦ã„ããŸã„ã¨æ€ã„ã¾ã™ã€‚ç•°å¸¸æ¤œçŸ¥ã€æƒ…å ±æ¤œç´¢ã€è‡ªç„¶è¨€èªžå‡¦ç†ã¨ã„ã£ãŸå¤šæ§˜ãªåˆ†é‡Žã®è«–æ–‡ãŒç™»å ´ã—ã€èªã¿ä¼šã‚‚ç››ã‚Šä¸ŠãŒã‚Šã¾ã—ãŸã€‚ãªãŠã€è«–æ–‡PDFã¯å„å¦ä¼šãƒšãƒ¼ã‚¸ã¾ãŸã¯è‘—è€…ã®Webã‚µã‚¤ãƒˆã§å…¬é–‹ã•ã‚Œã¦ã„ã‚‹ã‚‚ã®ã‚’å‚ç…§ã—ã¦ã„ã¾ã™ã€‚ Sparse Gaussian Markov Random Field Mixtures for Anomaly Detection Change Detection using Directional Statistics Linguistically Regularized LSTM for Sentiment Classification BitFunnel: Revi
syou6162 2017/09/27
ç•°å¸¸æ¤œçŸ¥ã€æƒ…å ±æ¤œç´¢ã€è‡ªç„¶è¨€èªžå‡¦ç†ã¨è‰²ã‚“ãªè©±é¡Œã§ç››ã‚Šä¸ŠãŒã‚Šã¾ã—ãŸ!

machine_learning

ml

nlp
ãƒªãƒ³ã‚¯
Tutorial on Variational Autoencodersã‚’èªã‚€ - ç·‘èŒ¶æ€è€ƒãƒ–ãƒã‚°
Variational Autoencodersï¼Œç•¥ã—ã¦VAEsã¨å‘¼ã°ã‚Œã‚‹ Arxiv: https://arxiv.org/abs/1606.05908 Tutorial on Variational Autoencoders(VAEs)ã‚’èªã¿è§£ã„ã¦ã„ã“ã†ã¨æ€ã†ã€‚ å…ˆäººãŸã¡ã«ã‚ˆã‚‹æ—¥æœ¬èªžã®è©³ç´°ãªè§£èª¬ã¯ãƒãƒƒãƒˆä¸Šã«ã‚´ãƒã‚´ãƒã‚ã‚‹ã®ã§ã€ æœ¬è¨˜äº‹ã¯è‡ªåˆ†è‡ªèº«ã®ç†è§£ã®ãŸã‚ã®ãƒ¡ãƒ¢ã¨ã„ã†ä½ç½®ã¥ã‘ã€‚ æ½œåœ¨å¤‰æ•° ç”Ÿæˆãƒ¢ãƒ‡ãƒ«ã¯ã€é«˜æ¬¡å…ƒç©ºé–“ã«ãŠã‘ã‚‹ãƒ‡ãƒ¼ã‚¿ã®åˆ†å¸ƒè‡ªä½“ã‚’ãƒ¢ãƒ‡ãƒ«åŒ–ã™ã‚‹ã‚‚ã®ã®ã“ã¨ã€‚ ä¾‹ãˆã°ã€MNISTã®ã‚ˆã†ãª0ã€œ9ã‹ã‚‰ãªã‚‹æ‰‹æ›¸ãæ–‡å—ã‚’ç”Ÿæˆã™ã‚‹ç›®çš„ã®ç”Ÿæˆãƒ¢ãƒ‡ãƒ«ãŒã‚ã‚‹ï¼Ž ã“ã®ã¨ãã€ãƒ¢ãƒ‡ãƒ«ã®ç”Ÿæˆå¯¾è±¡ã¯0ã€œ9ã®æ–‡å—ãªã®ã§ã€ ãƒ¢ãƒ‡ãƒ«ã¯ã¾ãšã©ã®æ–‡å—ã‚’ç”Ÿæˆã™ã‚‹ã‹ã‚’æ±ºã‚ã¦ã€ãã®æ±ºå®šã®ã‚‚ã¨ã§ãƒ‡ãƒ¼ã‚¿ã‚’ç”Ÿæˆã™ã‚‹ã‚‚ã®ã ï¼Œã¨ä»®å®šã—ã¦ã¿ã‚‹ã€‚ ã“ã®ã¨ãã®äºˆã‚æ±ºå®šã—ã¦ãŠããƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã®ã“ã¨ã‚’æ½œåœ¨å¤‰æ•°(latent variable) ã¨å‘¼ã¶
syou6162 2017/05/26
machine learning

ml
ãƒªãƒ³ã‚¯
CycleGAN
Transf erring characteristics from one image to another is an exciting proposition. How cool would it be if you could take a photo and convert it into the style of Van Gogh or Picasso! These are examples of cross domain image transfer - we want to take an image from an input domain $D_i$ and then transf orm it into an image of target domain $D_t$ without necessarily having a one-to-one mapping betwe
syou6162 2017/05/26
deep learning

ml

machine learning
ãƒªãƒ³ã‚¯
æ–‡å—èªè˜ã¯CNNã§çµ‚ã‚ã‚‹ã®ã‹ï¼Ÿ
é›»åæƒ…å ±é€šä¿¡å¦ä¼šã€Œãƒ‘ã‚¿ãƒ¼ãƒ³èªè˜ãƒ»ãƒ¡ãƒ‡ã‚£ã‚¢ç†è§£ç ”ç©¶ä¼šã€ï¼ˆ2016å¹´2æœˆ14æ—¥ï¼ ä¹å·žå·¥æ¥å¤§å¦ï¼Œç¦å²¡çœŒé£¯å¡šå¸‚)ã§ã®ãƒ—ãƒ¬ã‚¼ãƒ³è³‡æ–™ã§ã™ï¼Ž å¯¾å¿œã™ã‚‹åŽŸç¨¿ã¯ä»¥ä¸‹ã§ã™ï¼Ž é›»åæƒ…å ±é€šä¿¡å¦ä¼šæŠ€è¡“ç ”ç©¶å ±å‘Š, PRMU2015-133 http://www.ieice.org/ken/paper/20160221UbGo/ ä»¥ä¸‹ã¯ã‚¢ãƒ–ã‚¹ãƒˆãƒ©ã‚¯ãƒˆã§ã™ï¼Ž=========================== å°åˆ·æ•°å—ï¼Œæ‰‹æ›¸ãæ•°å—ï¼Œå¤šãƒ•ã‚©ãƒ³ãƒˆæ•°å—ã‚’å¯¾è±¡ã¨ã—ã¦ï¼Œç•³ã¿è¾¼ã¿ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒ ãƒˆãƒ¯ãƒ¼ã‚¯(CNN) ã«ã‚ˆã‚‹èªè˜å®Ÿé¨“ã‚’è©¦ã¿ãŸï¼Žã„ãšã‚Œã®ã‚¿ã‚¹ã‚¯ã«ã‚‚å¤§è¦æ¨¡ãª ãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã‚’ç”¨ã„ãŸï¼Žå¾—ã‚‰ã‚ŒãŸèªè˜çŽ‡ã¯ï¼Œå°åˆ·æ•°å—ã«ã¤ã„ã¦99.99%ï¼Œæ‰‹æ›¸ãæ•°å—ã«ã¤ã„ã¦99.89%ï¼Œãã—ã¦å¤šãƒ•ã‚©ãƒ³ãƒˆæ•°å—ã«ã¤ã„ã¦96.4%ã§ã‚ã£ãŸï¼Ž ã•ã‚‰ã«å°åˆ·æ•°å—ã¨æ‰‹æ›¸ãæ•°å—ã®æ··åˆèªè˜ã¨ã„ã†ï¼Œäºˆæƒ³ã•ã‚Œã‚‹å›°é›£æ€§ã‹ã‚‰ã‹å¾“æ¥ã‚ã¾ã‚Šè©¦ã¿ã‚‰ã‚Œãªã‹ã£ãŸèª²é¡Œã«ã¤ã„ã¦ã‚‚ï¼ŒCNNã®åˆ©
syou6162 2017/02/03
deep learning

ml
ãƒªãƒ³ã‚¯
æ©Ÿæ¢°å¦ç¿’ãƒ¢ãƒ‡ãƒ«ã«å¯¾ã™ã‚‹æ”»æ’ƒæ‰‹æ³• -Equation-Solving Attacks- | æŠ€è¡“è€…ãƒ–ãƒã‚° | ä¸‰äº•ç‰©ç”£ã‚»ã‚ãƒ¥ã‚¢ãƒ‡ã‚£ãƒ¬ã‚¯ã‚·ãƒ§ãƒ³æ ªå¼ä¼šç¤¾
ã“ã®1ï½ž2å¹´ã®é–“ã«æ©Ÿæ¢°å¦ç¿’ã‚’ä½¿ç”¨ã—ãŸã‚µãƒ¼ãƒ“ã‚¹ã‚„ãƒ—ãƒãƒ€ã‚¯ãƒˆãŒæ•°å¤šããƒªãƒªãƒ¼ã‚¹ã•ã‚Œã¦ã„ã¾ã™ã€‚ãã‚Œã¨åŒæ™‚ã«ã€æ©Ÿæ¢°å¦ç¿’ãƒ¢ãƒ‡ãƒ«ã«å¯¾ã™ã‚‹æ”»æ’ƒæ‰‹æ³•ã‚‚å¤šãç™ºè¡¨ã•ã‚Œã¦ã„ã¾ã™[1] [2] [3]ã€‚ æœ¬ãƒ–ãƒã‚°ã§ã¯ã€EPFLã®Florianæ°ã‚‰ãŒç™ºè¡¨ã—ãŸè«–æ–‡ã€ŒStealing Machine Learning Models via Prediction APIsã€ã‹ã‚‰ã€ŒModel Extraction Attacksã€ã¨å‘¼ã°ã‚Œã‚‹æ©Ÿæ¢°å¦ç¿’ãƒ¢ãƒ‡ãƒ«ã«å¯¾ã™ã‚‹æ”»æ’ƒæ‰‹æ³•ã®ä¸€ä¾‹ã‚’ç´¹ä»‹ã—ã¾ã™ã€‚ Model Extraction Attacks ã€ŒModel Extraction Attacksã€ã¨ã¯ã€æ©Ÿæ¢°å¦ç¿’ã‚µãƒ¼ãƒ“ã‚¹(ML Service)ã«å¯¾ã™ã‚‹å…¥åŠ›å€¤ã¨ã€ãã‚Œã«å¯¾å¿œã™ã‚‹å‡ºåŠ›å€¤ã‚’åŸºã«ã€ML Serviceã§ä½¿ç”¨ã—ã¦ã„ã‚‹MLãƒ¢ãƒ‡ãƒ«(Logistic Regressionã€Multilayer Perceptronã€Decision
syou6162 2017/01/20
ml

machinelearning
ãƒªãƒ³ã‚¯
site-reliability-engineering/6-DistributedMonitoringSystem.md at master Â· sysbooks/site-reliability-engineering
syou6162 2016/10/28
ml_study

ml
ãƒªãƒ³ã‚¯
ãƒ¡ãƒ³ãƒ†ãƒŠãƒ–ãƒ«ãªæ©Ÿæ¢°å¦ç¿’ã‚·ã‚¹ãƒ†ãƒ ã®è€ƒå¯Ÿ | æ ªå¼ä¼šç¤¾ã‚µã‚¤ãƒãƒ¼ã‚¨ãƒ¼ã‚¸ã‚§ãƒ³ãƒˆ
åŸ·ç†è€… åŸ·ç†è€…ï¼šæ•¸è¦‹æ‹“æœ— æ‰€å±žéƒ¨ç½²ï¼šæŠ€è¡“æœ¬éƒ¨ ç§‹è‘‰åŽŸãƒ©ãƒœ æ¥å‹™çµŒæ´ï¼šãƒˆãƒ¬ãƒ³ãƒ‰æ¤œçŸ¥ã‚·ã‚¹ãƒ†ãƒ ã‚„ã‚¹ãƒ‘ãƒ ãƒ•ã‚£ãƒ«ã‚¿ã®é–‹ç™ºãƒ»é‹ç”¨ã€Amebaã¨åºƒå‘Šé…ä¿¡ã®ãƒ‡ãƒ¼ã‚¿åˆ†æžæ¥å‹™ã‚’æ‹…å½“ã€‚ æ¦‚è¦ æ©Ÿæ¢°å¦ç¿’ã¯ã€ãƒ‡ãƒ¼ã‚¿ã®è¤‡é›‘ãªé–¢ä¿‚ã‚’æ˜Žã‚‰ã‹ã«ã—ã€æœªçŸ¥ãƒ‡ãƒ¼ã‚¿ã®äºˆæ¸¬ã‚„ç‰¹å¾´ã®ç™ºè¦‹ã‚’å¾—ã‚‹ã®ã«æœ‰ç”¨ã§ã‚ã‚‹ã€‚ã—ã‹ã—ã€ä¸€èˆ¬ã«æä¾›ã•ã‚Œã¦ã„ã‚‹æ©Ÿæ¢°å¦ç¿’ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã¯ã€é•·æœŸçš„ãªé‹ç”¨ã‚’ç›®æ¨™ã¨ã—ã¦ãƒ‡ã‚¶ã‚¤ãƒ³ã•ã‚Œã¦ã„ãªã„ã€‚ãã®ãŸã‚ã€é«˜ã„è˜åˆ¥æ€§èƒ½ã‚’ç¶æŒã™ã‚‹ã«ã¯ã€é€šå¸¸ã®ã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢é–‹ç™ºãƒ»é‹ç”¨ã®çŸ¥è¦‹ã§ã¯ä¸ååˆ†ã§ã‚ã‚‹ã€‚æœ¬ãƒ¬ãƒãƒ¼ãƒˆã®ç›®çš„ã¯ã€ã‚¢ãƒ¡ãƒ¼ãƒãƒ–ãƒã‚°ã«æä¾›ã—ã¦ã„ã‚‹æ©Ÿæ¢°å¦ç¿’ãƒ•ã‚£ãƒ«ã‚¿ã‚’ä¾‹ã¨ã—ã¦ã€é•·æœŸçš„ã«é‹ç”¨å¯èƒ½ãªæ©Ÿæ¢°å¦ç¿’ã‚·ã‚¹ãƒ†ãƒ ã‚’è€ƒå¯Ÿã™ã‚‹ã“ã¨ã§ã‚ã‚‹ã€‚ ç›®æ¬¡ 1.ã¯ã˜ã‚ã« ã‚¢ãƒ¡ãƒ–ãƒã¯ï¼Œã‚¢ãƒ¡ãƒ¼ãƒã§æœ€ã‚‚å¤ãï¼Œæƒ…å ±ç™ºä¿¡æ‰‹æ®µã¨ã—ã¦å¤šãã®ãƒ¦ãƒ¼ã‚¶ã«åˆ©ç”¨ã•ã‚Œã¦ã„ã‚‹ãƒ–ãƒã‚°ã‚µãƒ¼ãƒ“ã‚¹ã§ã‚ã‚‹ï¼Ž2015å¹´2æœˆã®ãƒ‡ãƒ¼ã‚¿ã§ã¯ï¼Œä¸€æ—¥å½“ãŸã‚Šç´„æ•°åä¸‡ä»¶ã®ã‚¨ãƒ³ãƒˆãƒªãƒ¼ãŒæŠ•ç¨¿ã•ã‚Œã¦ã„ã‚‹ï¼Žä¸€æ–¹ã§ï¼Œä¸€éƒ¨ã®æ‚ªæ„
syou6162 2016/10/04
ml_study

ml
ãƒªãƒ³ã‚¯
number.of.queries.eps
A tutorial on active learning Sanjoy Dasgupta1 John Langford2 UC San Diego1 Yahoo Labs2 Exploiting unlabeled data A lot of unlabeled data is plentiful and cheap, eg. documents off the web speech samples images and video But labeling can be expensive. Exploiting unlabeled data A lot of unlabeled data is plentiful and cheap, eg. documents off the web speech samples images and video But labeling can
syou6162 2016/09/23
active learning

ml
ãƒªãƒ³ã‚¯
fp1022-chu.dvi
Unbiased Online Active Learning in Data Streams Wei Chu âˆ— Microsoft Redmond, WA, USA chu.wei@microsoft.com Martin Zinkevich Yahoo! Labs Sunnyvale, CA, USA maz@yahoo-inc.com Lihong Li Yahoo! Labs Sunnyvale, CA, USA lihong@yahoo-inc.com Achint Thomas Yahoo! Labs Sunnyvale, CA, USA aothomas@yahoo-inc.com Belle Tseng Yahoo! Labs Sunnyvale, CA, USA belle@yahoo-inc.com ABSTRACT Unlabeled samples can be
syou6162 2016/09/23
ml

active learning
ãƒªãƒ³ã‚¯
Machine Learning: The High Interest Credit Card of Technical Debt
Philosophy We strive to create an environment conducive to many different types of research across many different time scales and levels of risk. Learn more about our Philosophy Learn more
syou6162 2016/07/21
ML
ãƒªãƒ³ã‚¯
Anomaly Monitor
Product { this.openCategory = category; const productMenu = document.querySelector('.product-menu'); window.DD_RUM.onReady(function() { if (productMenu.classList.contains('show')) { window.DD_RUM.addAction(`Product Category ${category} Hover`) } }) }, 160); }, clearCategory() { clearTimeout(this.timeoutID); } }" x-init=" const menu = document.querySelector('.product-menu'); var observer = new Muta
syou6162 2016/07/21
datadog

ml
ãƒªãƒ³ã‚¯
ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã«ã‚ˆã‚‹ è‡ªç„¶è¨€èªžå‡¦ç† (æŠ€è¡“ç·¨)
åå¤å±‹å¤§å¦ç‰¹åˆ¥è¬›ç¾© 2016å¹´6æœˆ29æ—¥(æ°´) ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã«ã‚ˆã‚‹ è‡ªç„¶è¨€èªžå‡¦ç† (æŠ€è¡“ç·¨) æ—¥æœ¬ã‚¢ã‚¤ãƒ»ãƒ“ãƒ¼ãƒ»ã‚¨ãƒ æ ªå¼ä¼šç¤¾ æ±äº¬åŸºç¤Žç ”ç©¶æ‰€ åªäº• ç¥å¤ª [email protected] 1 ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯æŠ€è¡“è©³ç´° â€¢ ç›®çš„é–¢æ•° â€¢ èª¤å·®é–¢æ•° â€¢ ç›®çš„é–¢æ•°ã®æœ€å°åŒ– â€¢ å‹¾é…æ³• â€¢ ç›®çš„é–¢æ•°ã®å¾®åˆ†è¨ˆç®— â€¢ èª¤å·®é€†ä¼æ¬æ³• â€¢ èª¤å·®ã®åˆ†è§£ã¨å¯¾å‡¦æ‰‹æ³• â€¢ æŽ¨å®šèª¤å·®ã«åŠ¹ãæ‰‹æ³• â€¢ æœ€é©åŒ–èª¤å·®ã«åŠ¹ãæ‰‹æ³• â€¢ RNNã®è©±é¡Œ 2 ç›®çš„é–¢æ•°: èª¤å·® â€¢ æ•™å¸«ã‚ã‚Šå¦ç¿’ã®ç›®çš„é–¢æ•° â€¢ ð‘¥ âˆˆ ð‘‹: å…¥åŠ›, ð‘¦ âˆˆ ð‘Œ: å‡ºåŠ› â€¢ å…¥åŠ›xã‹ã‚‰yã‚’äºˆæ¸¬ã—ãŸã„å•é¡Œè¨å®š â€¢ çœŸã®ç›®çš„é–¢æ•°: ð¿âˆ— ðœƒ = âˆ«ð‘ ð‘¥, ð‘¦ â„“ðœƒ ð‘¥, ð‘¦ ð‘‘ð‘¥ð‘‘ð‘¦ â€¢ â„“ðœƒã¯äº‹ä¾‹ã”ã¨ã®æå¤±é–¢æ•°ï¼ˆå¾Œè¿°ï¼‰ â€¢ è¨“ç·´ãƒ‡ãƒ¼ã‚¿ã§ã®èª¤å·® â€¢ ãƒ‡ãƒ¼ã‚¿åˆ†å¸ƒp(x,y)ã¯æ™®é€šã‚ã‹ã‚‰ãªã„ã®ã§è¨“ç·´ãƒ‡ãƒ¼ã‚¿Nå€‹: D=
syou6162 2016/07/01
dnn

ml
ãƒªãƒ³ã‚¯
æ·±å±¤ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ï¿½ã«ã‚ˆã‚‹çŸ¥è˜ã®è‡ªå‹•ç²å¾—ãƒ»æŽ¨è«–
2. è‡ªç„¶è¨€èªžå‡¦ç†ã¨ã¯ â€¢ è¨€è‘‰ã‚’æ“ã‚‹è³¢ã„ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ã‚’ä½œã‚‹ â€¢ å¿œç”¨: æƒ…å ±æ¤œç´¢ï¼Œæ©Ÿæ¢°ç¿»è¨³ï¼Œè³ªå•å¿œç”ï¼Œè‡ªå‹•è¦ ç´„ï¼Œå¯¾è©±ç”Ÿæˆï¼Œè©•åˆ¤åˆ†æžï¼ŒSNSåˆ†æžï¼Œâ€¦ â€¢ åŸºç¤Ž: å“è©žã‚¿ã‚°ä»˜ã‘ï¼ˆå½¢æ…‹ç´ è§£æžï¼‰ï¼Œãƒãƒ£ãƒ³ã‚ ãƒ³ã‚°ï¼Œå›ºæœ‰è¡¨ç¾æŠ½å‡ºï¼Œæ§‹æ–‡è§£æžï¼Œå…±å‚ç…§è§£æžï¼Œ æ„å‘³å½¹å‰²ä»˜ä¸Žï¼Œâ€¦ â€¢ å¤šãã®ã‚¿ã‚¹ã‚¯ã¯ã€Œå…¥åŠ›ð‘¥ð‘¥ã‹ã‚‰å‡ºåŠ›ï¿½ð‘¦ð‘¦ã‚’äºˆæ¸¬ã€ ï¿½ð‘¦ð‘¦ = argmax ð‘¦ð‘¦âˆˆð‘Œð‘Œ ð‘ƒð‘ƒ(ð‘¦ð‘¦|ð‘¥ð‘¥) â€»ç¢ºçŽ‡ã§ã¯ãªã„ãƒ¢ãƒ‡ãƒ«ã‚‚ã‚ã‚Šã¾ã™ 2016-06-30 æ·±å±¤ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã«ã‚ˆã‚‹çŸ¥è˜ã®è‡ªå‹•ç²å¾—ãƒ»æŽ¨è«– 2 3. å˜èªžåˆ—ã‹ã‚‰ãƒ©ãƒ™ãƒ«: Ì‚ð‘¦ð‘¦ = argmax ð‘¦ð‘¦âˆˆð‘Œð‘Œ ð‘ƒð‘ƒ(ð‘¦ð‘¦|ð’™ð’™) 2016-06-30 æ·±å±¤ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã«ã‚ˆã‚‹çŸ¥è˜ã®è‡ªå‹•ç²å¾—ãƒ»æŽ¨è«– 3 The movie is the best Iâ€™ve ever seen! The movie
syou6162 2016/06/30
ã‚ˆã•ã—ã‹ãªã„

nlp

ml
ãƒªãƒ³ã‚¯
Index of /lectures
Index of /lectures NameLast modifiedSizeDescription Parent DirectoryÂ - CS224D-Lecture7-2.pdf2016-04-19 16:06 77K CS224d-Lecture1.pdf2016-03-31 12:03 11M CS224d-Lecture2.pdf2016-03-31 12:03 5.0M CS224d-Lecture3.pdf2016-04-05 15:18 6.2M CS224d-Lecture 4.pdf2016-04-05 15:21 2.7M CS224d-Lecture4.pdf2016-04-07 13:16 2.6M CS224d-Lecture5.pdf2016-04-12 14:46 2.9M CS224d-Lecture6.pdf2016-04-17 18:52 4.2M
syou6162 2016/05/31
é¢ç™½ãã†

ml
ãƒªãƒ³ã‚¯
https://cs224d.stanford.edu/lectures/CS224d-Lecture7.pdf
syou6162 2016/05/31
tensorflow

python

ml
ãƒªãƒ³ã‚¯
TensorFlow å¤§é‡ã®ç”»åƒã‹ã‚‰å¦ç¿’ã™ã‚‹ã«ã¯ãƒ»ãƒ»ãƒ»ã€œ(ã»ã¼)è§£æ±ºç·¨ã€œ - Qiita
ã¯ã˜ã‚ã« ã©ã†ã‚‚ã€ikkiã§ã™ã€‚ã‚ˆã†ã‚„ãã€å¤§é‡ã®ç”»åƒã‚’TensorFlowã«é£Ÿã‚ã›ã‚‹ã®ã‹ã€è§£æ±ºã—ã¾ã—ãŸã€‚ è§£æ±ºã—ã¦ã¿ã‚Œã°ã€ãƒ›ãƒ³ãƒˆæ¥ãšã‹ã—ã„å†…å®¹ã§ã—ãŸãŒã€‚ã€‚ã€‚ ä¸é€”åŠç«¯ã«ç†è§£ã—ã¦ã®ã‚³ãƒ”ãƒšã¯ãƒ€ãƒ¡ã‚¼ãƒƒã‚¿ã‚¤ï¼ ä»¥å‰ã¾ã§ã®å•é¡Œç‚¹ ã“ã¡ã‚‰ã§è‡ªå‰ã§ç”¨æ„ã—ãŸç”»åƒã‚’å¦ç¿’ã™ã‚‹ãƒ—ãƒã‚°ãƒ©ãƒ ã‚’æŽ²è¼‰ã—ãŸãŒã€ã“ã®ã¾ã¾ã ã¨ã€æ•°ä¸‡æžšã®ç”»åƒãƒ‡ãƒ¼ã‚¿ã‚’å¦ç¿’ã™ã‚‹ã“ã¨ã¯ã§ããªã‹ã£ãŸã€‚ ã¨ã„ã†ã‚ã‘ã§ã€ã¾ãšã¯ä¿®æ£ã—ãŸãƒ—ãƒã‚°ãƒ©ãƒ ã‚’æŽ²è¼‰ã™ã‚‹ã€‚ #!/usr/bin/env python # -*- coding: utf-8 -*- import sys import cv2 import numpy as np import tensorflow as tf import tensorflow.python.platform NUM_CLASSES = 3 IMAGE_SIZE = 28 IMAGE_PIXELS = IMAGE_SIZE
syou6162 2016/05/31
tensorflow

python

ml
ãƒªãƒ³ã‚¯
Low Dimension Embeddings for Visualization
Representation learning is a hot area in machine learning. In natural language processing (for example), learning long vectors for words has proven quite effective on several tasks. Often, these representations have several hundred dimensions. To perform a qualitative analysis of the learned representations, it helps to visualize them. Thus, we need a principled approach to drop from this high dim
syou6162 2014/11/01
clojure

ml
ãƒªãƒ³ã‚¯
Welcome to MLSS 2014
Machine Learning Summer School 2014 in Pittsburgh http://www.mlss2014.com See the website for more videos and slides. Welcome notes
syou6162 2014/07/26
ml

mlss
ãƒªãƒ³ã‚¯
The Department of Computer Science - Home - New
syou6162 2014/06/24
qpã‚’è§£ãpythonã‚³ãƒ¼ãƒ‰

nlp

python

mira

ml
ãƒªãƒ³ã‚¯
1 2 æ¬¡ã®ãƒšãƒ¼ã‚¸