id:nokuno ãã主å¬ã® NIPS èªã¿ä¼ã«ã®ãã®ãåå ããããã¨ããããã¾ãï¼ãç²ãæ§ã§ããï¼åä½
[Karger+ NIPS11] Iterative Learning for Reliable Crowdsourcing Systems ã¨ããã¯ã©ã¦ãã½ã¼ã·ã³ã°ãç¹ã« Amazon Mechanical Turk ã«ã¦ spammer ãããå ´åã«ãå¹æçãªæ¨è«ææ³ãææ¡ãã¤ã¤ãã¨ã©ã¼çã®ä¸éãè©ä¾¡ãããã¨ãããã®ã
å¿
ãæ£è§£ãæã¤ï¼å¤åé¡åé¡ã§ãworkers ãæ£è§£ãçãã確ç㯠tasks ã«å¯¾ãã¦ç¬ç«ã«æ±ºã¾ããã¨ããããªãå¼·ãä»®å®ãå
¥ã£ã¦ããã®ã¯ã¾ãè¯ãã®ã ããã¨ã©ã¼çã®ä¸éã®è©ä¾¡å¼ã®ãã©ã¡ã¼ã¿ã«ã worker ã®æ£è§£çã®ï¼æ¬¡ã¢ã¼ã¡ã³ããã¨ããçã®å¤ãç¥ãããããªãå¤ãå
¥ã£ã¡ãã£ã¦ãã¨ãããå人çã«ã¯ä»ã²ã¨ã¤ã
ã¾ãããããªãã¦ãããªãçãä¸é(å®é¨ã§ç¢ºèªã§ããã¨ã©ã¼çã¯ãã®ä¸éããã¯ããã«å°ãã)ãªã®ã§ããã®æã® boundary ã®ã話ã®å¸¸ã§ãããã¨ããããä¸çå¼ï¼åã§ãããã°ããããå®å¿ã ããï¼ãã¨ãããã¨ãªã®ããªãã¨ã
ä»åã¯持橋さんが参加してくださっててã豪è¯å
å®ã
ä»åãã¾ãæºåã®æéããªãã¦ãå½åã¯æå
ã(ç¬)ãããã¯ã¢ãã«ãªè«æãèªããã¨ãã¦ããã®ã ãã©ãããã¨ãã¦è¯ãã£ãã¨ããæ°åååãããææ©ããã«ãã·ãã·çªã£è¾¼ãã§ããããæ©ä¼ãªãã¦ãªããªããªãããã¾ã£ããªããªæ°åååã
è³æã®ãã¡å ¬éããã¦ããåãçºè¡¨é ãæ¬ç§°ç¥ã
- [Shindler+] "Fast and Accurate K-means for Large Datasets" by @nokuno
- [Karger+] "Iterative Learning for Reliable Crowdsourcing Systems" by @shuyo
- [Leibo+] "Why The Brain Separates Face Recognition From Object Recognition" by @takmin
- [Yue+] "Linear submodular bandits and their application to diversified retrieval" by @tsubosaka
- [Ren+] "The Kernel Beta Process" by daiti-m
- [Recht+] "HOGWILD!: A Lock-Free Approach to Parallelizing Stochastic Gradient Descent" by @sleepy_yoshi
- [Chierichetti+] "Reconstructing Patterns of Information Discussion from Incomplete Observations" by @smly
- [Hernández-Lobato+] Robust Multi-Class Gaussian Process Classification" by @suzuvie_re