Dive into Deep Learning with 15 free online courses
By David Venturi Every day brings new headlines for how deep learning is changing the world around us. A few examples: Deep learning algorithm diagnoses skin cancer as well as seasoned dermatologists Amazon Go: How Deep Learning and AI will change Retailing Deep Learning Makes Driverless Cars Better at Spotting Pedestrians Want to see what the fuss is all about? Looking to master the technical con
飯塚里志 — ディープãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã«ã‚ˆã‚‹ç”»åƒè£œå®Œï¼ˆSIGGRAPH 2017）
概è¦ï¼š æœ¬ç ”ç©¶ã§ã¯ï¼Œç•³ã¿è¾¼ã¿ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’用ã„ã¦ï¼Œã‚·ãƒ¼ãƒ³ã®å¤§åŸŸçš„ã‹ã¤å±€æ‰€çš„ãªæ•´åˆæ€§ã‚’考慮ã—ãŸç”»åƒè£œå®Œã‚’è¡Œã†æ‰‹æ³•ã‚’æ案ã™ã‚‹ï¼Žæ案ã™ã‚‹è£œå®Œãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã¯å…¨å±¤ãŒç•³ã¿è¾¼ã¿å±¤ã§æ§‹æˆã•ã‚Œï¼Œä»»æ„ã®ã‚µã‚¤ã‚ºã®ç”»åƒã«ãŠã‘る自由ãªå½¢çŠ¶ã®ã€Œç©´ã€ã‚’補完ã§ãる．ã“ã®è£œå®Œãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã«ï¼Œã‚·ãƒ¼ãƒ³ã®æ•´åˆæ€§ã‚’考慮ã—ãŸç”»åƒè£œå®Œã‚’å¦ç¿’ã•ã›ã‚‹ãŸã‚，本物ã®ç”»åƒã¨è£œå®Œã•ã‚ŒãŸç”»åƒã‚’è˜åˆ¥ã™ã‚‹ãŸã‚ã®å¤§åŸŸè˜åˆ¥ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã¨å±€æ‰€è˜åˆ¥ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’構築ã™ã‚‹ï¼Žå¤§åŸŸè˜åˆ¥ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã¯ç”»åƒå…¨ä½“ãŒè‡ªç„¶ãªç”»åƒã«ãªã£ã¦ã„ã‚‹ã‹ã‚’評価ã—，局所è˜åˆ¥ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã¯è£œå®Œé ˜åŸŸå‘¨è¾ºã®ã‚ˆã‚Šè©³ç´°ãªæ•´åˆæ€§ã«ã‚ˆã£ã¦ç”»åƒã‚’評価ã™ã‚‹ï¼Žã“ã®2ã¤ã®è˜åˆ¥ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ä¸¡æ–¹ã‚’「ã ã¾ã™ã€ã‚ˆã†ã«è£œå®Œãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’å¦ç¿’ã•ã›ã‚‹ã“ã¨ã§ï¼Œã‚·ãƒ¼ãƒ³å…¨ä½“ã§æ•´åˆæ€§ãŒå–ã‚Œã¦ãŠã‚Šï¼Œã‹ã¤å±€æ‰€çš„ã«ã‚‚自然ãªè£œå®Œç”»åƒã‚’出力ã™ã‚‹ã“ã¨ãŒã§ãる．æ案手法ã«ã‚ˆã‚Šï¼Œæ§˜ã€…ãªã‚·ãƒ¼ãƒ³ã«ãŠã„ã¦è‡ªç„¶ãªç”»åƒè£œå®ŒãŒå¯èƒ½ã¨ãªã‚Šï¼Œã•ã‚‰ã«å¾“æ¥ã®
When DNNs go wrong – adversarial examples and what we can learn from them | the morning paper
the morning paper a random walk through Computer Science research, by Adrian Colyer Made delightfully fast by strattic Yesterday we looked at a series of papers on DNN understanding, generalisation, and transfer learning. One additional way of understanding what’s going on inside a network is to understand what can break it. Adversarial examples are deliberately constructed inputs which cause a ne
Why does deep and cheap learning work so well?
We show how the success of deep learning could depend not only on mathematics but also on physics: although well-known mathematical theorems guarantee that neural networks can approximate arbitrary functions well, the class of functions of practical interest can frequently be approximated through "cheap learning" with exponentially fewer parameters than generic ones. We explore how properties freq
📠橘ã‚ã‚Šã™åˆ†é¡žå™¨ - ã¿ãšã´ãƒ¼æ—¥è¨˜
GWã«æ—¥å…‰ã«é–‹ç™ºåˆå®¿ã«è¡Œã£ã¦ããŸã€‚ 「機械å¦ç¿’ã§éŠã¼ã†ã€ã‚’テーマã«ã—ãŸã®ã§ã€æ©˜ã‚ã‚Šã™åˆ†é¡žå™¨ã‚’@banjunã¨ä½œæˆã—ãŸã€‚ æˆæžœç‰© m@gic-with-arisu from MIZUNO Hiroki on Vimeo. ã‚ã‚Šã™ã«ã¯é’æž ã‚’ã€ãれ以外ã«ã¯èµ¤æž を表示ã—ã¦ã„る。 宿 日光æ±ç…§å®®è‡³è¿‘ã®ç‚ç«ç„¼ãã®å®¿ãƒ»ãƒšãƒ³ã‚·ãƒ§ãƒ³ã¯ã˜ã‚ã®ã„ã£ã½ã«å®¿æ³Šã—ãŸã€‚開発åˆå®¿ã«ç†è§£ã®ã‚る宿ã§ã‚ˆã‹ã£ãŸã€‚ãŸã ã€ç†è§£ãŒã‚ã‚‹ã‚ã¾ã‚Šã€Œæ˜¨æ—¥ã¯æˆæžœã§ã¾ã—ãŸã‹?ã€ã¨èžã‹ã‚Œã¦ã¤ã‚‰ã„æ€ã„ã‚‚ã—ãŸã€‚ 進æ—ã‚’èžã‹ã‚Œã‚‹ã‚ˆã‚Šã€æˆæžœã‚’èžã‹ã‚Œã‚‹ã»ã†ãŒåŽ³ã—ã„。 ã¾ãŸæ—§åž‹ã®iMacãŒã‚ã£ã¦ã‚ˆã‹ã£ãŸã€‚èµ·å‹•ã—ãŸã‚‰ãƒãƒ¼ã‚¸ãƒ§ãƒ³1ã®Safariã‚„ã€Mac版ã®IEãŒå…¥ã£ã¦ã¦æ¥½ã—ã‹ã£ãŸã€‚ ã‚ã‚Šã™ã®èª¿é” @banjunãŒãƒ‡ãƒ¬ã‚¹ãƒ†ã®ã‚¬ã‚·ãƒ£ã‚’引ãã¾ãã£ã¦ã€SSRã‚ã‚Šã™ã‚’調é”ã—ãŸã€‚ é™å®šSSRã¨ã„ã†ã®ã¯ï¼Œã‚„ã°ã„ (ã„ã¾ã“ã“)— ã°ã‚“ã˜ã‚…ã‚“(!!) (@banjun) 2016
サービス終了ã®ãŠçŸ¥ã‚‰ã›
サービス終了ã®ãŠçŸ¥ã‚‰ã› ã„ã¤ã‚‚Yahoo! JAPANã®ã‚µãƒ¼ãƒ“スをã”利用ã„ãŸã ãèª ã«ã‚ã‚ŠãŒã¨ã†ã”ã–ã„ã¾ã™ã€‚ ãŠå®¢æ§˜ãŒã‚¢ã‚¯ã‚»ã‚¹ã•ã‚ŒãŸã‚µãƒ¼ãƒ“スã¯æœ¬æ—¥ã¾ã§ã«ã‚µãƒ¼ãƒ“スを終了ã„ãŸã—ã¾ã—ãŸã€‚ 今後ã¨ã‚‚Yahoo! JAPANã®ã‚µãƒ¼ãƒ“スをã”愛顧ãã ã•ã„ã¾ã™ã‚ˆã†ã€ã‚ˆã‚ã—ããŠé¡˜ã„ã„ãŸã—ã¾ã™ã€‚
åŒå¯¾åˆ†è§£ã«ã‚ˆã‚‹æ§‹é€ å¦ç¿’ - Preferred Networks Research & Development
入力\(x\)ã‹ã‚‰å‡ºåŠ›\(y\)ã¸ã®é–¢æ•°ã‚’å¦ç¿’ã™ã‚‹æ©Ÿæ¢°å¦ç¿’ã®ä¸ã§ã€å‡ºåŠ›ãŒæ§‹é€ を有ã—ã¦ã„ã‚‹å•é¡Œã¯æ§‹é€ å¦ç¿’（Structured Output Learning）ã¨å‘¼ã°ã‚Œã€è‡ªç„¶è¨€èªžå‡¦ç†ã‚’ã¯ã˜ã‚ã€æ¤œç´¢ã®ãƒ©ãƒ³ã‚ングå¦ç¿’ã€ç”»åƒè§£æžã€è¡Œå‹•åˆ†æžãªã©å¤šãã®åˆ†é‡Žã§ã¿ã‚‰ã‚Œã¾ã™ã€‚ 今回ã¯ãã®ä¸ã§ã‚‚複数ã®æ§‹é€ æƒ…å ±ã‚’çµ„ã¿åˆã‚ã›ã¦ã‚‚効率的ã«å¦ç¿’・推論ãŒã§ãã‚‹åŒå¯¾åˆ†è§£ã«ã‚ˆã‚‹æ§‹é€ å¦ç¿’ã«ã¤ã„ã¦ç´¹ä»‹ã‚’ã—ã¾ã™ã€‚ # æ§‹é€ å¦ç¿’ã«ã¤ã„ã¦ã‚ˆã知ã£ã¦ã„ã‚‹ã¨ã„ã†æ–¹ã¯åŒå¯¾åˆ†è§£ã«ã‚ˆã‚‹æ§‹é€ å¦ç¿’ã®ã¨ã“ã‚ã¾ã§èªã¿é£›ã°ã—ã¦ãã ã•ã„。 æ§‹é€ å¦ç¿’ã®å°Žå…¥ æ§‹é€ ã‚’æœ‰ã—ãŸå‡ºåŠ›ã®ä¾‹ã¨ã—ã¦ã€ ラベル列　（å“è©žã€å½¢æ…‹ç´ 列ã®æŽ¨å®šã€æ™‚系列ã«ãŠã‘るアクションã®æŽ¨å®šã€ã‚»ãƒ³ã‚µåˆ—） 木　　　　（係りå—ã‘解æžã«ãŠã‘ã‚‹ä¿‚ã‚Šå—ã‘木ã€æ§‹æ–‡è§£æžæœ¨ã€è«‡è©±åˆ†æžã€å› 果分æžï¼‰ グラフ　　（DAG:è¿°èªžé …æ§‹é€ ã«ã‚ˆã‚‹æ„味解æžã€€äºŒéƒ¨ã‚°ãƒ©ãƒ•ãƒžãƒƒãƒãƒ³ã‚°:機械翻訳ã®å˜èªžå¯¾å¿œï¼‰ é †ä½ä»˜é›†åˆï¼ˆæ¤œç´¢ã«ãŠã‘ã‚‹é †ä½
1
ランã‚ング
ランã‚ング
ランã‚ング
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
Machine Learning | Coursera
Kaggle: The Home of Data Science
{{#tags}}- {{label}}
{{/tags}}