ðarXivã«ããã競銬ãã«é¢é£ããããªè«æð
â»å®å šã«è¶£å³ã®é åã§ã
æè¿ã¯AIã»LLMé¢é£ã®è«æãæ¥åžžçã«é²èŠ§ããããšãå€ããªã£ãŠãããããµãšæãç«ã£ãŠç«¶éŠ¬ã«é¢é£ããããªè«æãæ¢ããŠã¿ãã
â»è«æã¯Claude-3.5-Sonnetã§èªãŸããŠæ°ã«ãªã£ããQAããŠãŸãã
â»é äžåãçµæ§å€ããã®ããã
Efficient Market Dynamics: Unraveling Informational Efficiency in UK Horse Racing Betting Markets Through Betfair's Time Series Analysis
以äžClaude-3.5-Sonnetã«ããèœåããã³ãã
å¹ççåžå Žã®ãã€ããã¯ã¹ïŒBetfairã®æç³»ååæãéããè±åœç«¶éŠ¬ãããã£ã³ã°åžå Žã«ãããæ å ±å¹çæ§ã®è§£æ
ã©ããªãã®ïŒ
ãã®è«æã¯ãè±åœã®ç«¶éŠ¬åžå Žã«ããããããã£ã³ã°ååŒæBetfairã®ããŒã¿ãåæãããªã³ã©ã€ã³ã»ã¹ããŒãã»ãããã£ã³ã°ååŒæã®ç¹æ§ãæããã«ãããã®ã§ããå ·äœçã«ã¯ãéèåžå Žã§èŠ³å¯ããããå®ååãããäºå®ããšåŒã°ããçµ±èšçç¹åŸŽãããããã£ã³ã°ååŒæã®ããŒã¿ã«é©çšããŠèª¿æ»ããŠããŸãã
åæã®çµæããããã£ã³ã°ååŒæã®ãªã¿ãŒã³ååžã«ã¯ã軜ã裟ããèŠãããéèåžå Žã§äžè¬çã«èŠ³å¯ããããéã裟ããšã¯å¯Ÿç §çã§ããããšãåãããŸããããŸããå©åŸãšæ倱ã®é察称æ§ã¯èŠãããããªã¿ãŒã³ã¯å¹³åååž°æ§ãšå®åžžæ§ã瀺ããŸãããããã«ãèªå·±çžé¢ã®æžè¡°ãéãããšãæããã«ãªãããããã®ç¹åŸŽã¯ãããã£ã³ã°ååŒæãé«ãæ å ±å¹çæ§ãæã€å¯èœæ§ã瀺åããŠããŸãã
ãã®ç 究ã¯ããããã£ã³ã°ååŒæãšéèåžå Žã®é¡äŒŒç¹ãšçžéç¹ãæããã«ãããªã³ã©ã€ã³ã»ã¹ããŒãã»ãããã£ã³ã°ååŒæã®ç¬èªã®çµ±èšçç¹æ§ãç解ããããã®åºç€ãæäŸããŠããŸãã
å è¡ç 究ãæ¯ã¹ãŠã©ãããããïŒ
ãã®ç 究ã®é©æ°çãªç¹ã¯ãéèåžå Žã§é·å¹Žç 究ãããŠãããå®ååãããäºå®ãã®æŠå¿µããæ¯èŒçæ°ããåéã§ãããªã³ã©ã€ã³ã»ã¹ããŒãã»ãããã£ã³ã°ååŒæã«é©çšããããšã§ããå è¡ç 究ã§ã¯ããããã£ã³ã°ååŒæã®å¹çæ§ãäºæž¬ç²ŸåºŠã«çŠç¹ãåœãŠããã®ãå€ãã£ãã®ã«å¯Ÿãããã®ç 究ã¯ååŒæã®ããŒã¿ã®çµ±èšçç¹æ§ãå æ¬çã«åæããŠããŸãã
ç¹ã«ããªã¿ãŒã³ååžã®è£Ÿã®æ§è³ªãå©åŸæ倱ã®é察称æ§ãæéäŸåæ§ãªã©ã®ç¹åŸŽã詳现ã«èª¿æ»ããŠããŸããããã«ããããããã£ã³ã°ååŒæãéèåžå Žãšã¯ç°ãªãç¹æ§ãæã€ããšãæããã«ããŸãããäŸãã°ãéèåžå Žã§ã¯äžè¬çã«èŠ³å¯ããããéã裟ããããããã£ã³ã°ååŒæã§ã¯èŠãããªãããšã瀺ããŠããŸãã
ãŸãããã®ç 究ã¯å€§èŠæš¡ãªBetfairã®ããŒã¿ã»ããïŒ100äžä»¥äžã®äŸ¡æ Œå€åã·ã°ãã«ã73ã®åžå Žã10ã®ã€ãã³ãïŒã䜿çšããŠãããå è¡ç 究ãããå æ¬çãªåæãå¯èœã«ããŠããŸããããã«ããããããã£ã³ã°ååŒæã®ç¹æ§ãããæ£ç¢ºã«ææ¡ããéèåžå Žãšã®æ¯èŒãè¡ãããšãã§ããŠããŸãã
æè¡ãææ³ã®èã¯ã©ãïŒ
ãã®ç 究ã®äž»èŠãªæè¡çã¢ãããŒãã¯ãéèåžå Žã®åæã§äœ¿çšãããçµ±èšçææ³ããããã£ã³ã°ååŒæã®ããŒã¿ã«é©çšããããšã§ããå ·äœçã«ã¯ä»¥äžã®ææ³ãçšããããŠããŸãïŒ
- ãªã¿ãŒã³ååžã®åæïŒå¯Ÿæ°ãªã¿ãŒã³ãèšç®ãããã®çµ±èšçç¹æ§ïŒå¹³åãæšæºåå·®ãæªåºŠãå°åºŠïŒã調æ»ããŠããŸãã
- 裟ã®éãã®åæïŒHillæšå®éã䜿çšããŠååžã®è£Ÿã®ææ°ãæšå®ããäžè¬åã¬ãŠã¹ååžãããŒã¿ã«ãã£ããã£ã³ã°ããŠããŸãã
- å©åŸæ倱ã®é察称æ§ã®åæïŒæ£ãšè² ã®ãªã¿ãŒã³ã®ååžãæ¯èŒããã³ã«ã¢ãŽããã»ã¹ãã«ããæ€å®ãå®æœããŠããŸãã
- æéäŸåæ§ã®åæïŒèªå·±çžé¢é¢æ°ãèšç®ãããã®æžè¡°ã®éãã調ã¹ãŠããŸãããŸãã絶察ãªã¿ãŒã³ã®éç·åœ¢èªå·±çžé¢ãåæããŠããŸãã
- å®åžžæ§ã®æ€å®ïŒADFæ€å®ãšKPSSæ€å®ã䜿çšããŠããŒã¿ã®å®åžžæ§ã確èªããŠããŸãã
ãããã®ææ³ãçµã¿åãããããšã§ããããã£ã³ã°ååŒæã®ããŒã¿ã®å€é¢çãªç¹æ§ãæããã«ããéèåžå Žãšã®æ¯èŒãå¯èœã«ããŠããŸãããŸããããŒã¿ã®ååŠçããåžå Žããšã®åå¥åæãªã©ã詳现ãªåæææ³ãç 究ã®éèŠãªéšåãšãªã£ãŠããŸãã
ã©ããã£ãŠæå¹ã ãšæ€èšŒããïŒ
ãã®ç 究ã®æå¹æ§ã¯ãäž»ã«ä»¥äžã®æ¹æ³ã§æ€èšŒãããŠããŸãïŒ
倧èŠæš¡ããŒã¿ã»ããã®äœ¿çšïŒ100äžä»¥äžã®äŸ¡æ Œå€åã·ã°ãã«ã73ã®åžå Žã10ã®ã€ãã³ããå«ãå æ¬çãªããŒã¿ã»ãããåæããããšã§ãçµæã®ä¿¡é Œæ§ãé«ããŠããŸãã
æ¢åã®éèåžå Žç 究ãšã®æ¯èŒïŒãªã¿ãŒã³ååžãèªå·±çžé¢ããã©ãã£ãªãã£ã»ã¯ã©ã¹ã¿ãªã³ã°ãªã©ãéèåžå Žã§ well-established ãªãå®ååãããäºå®ããšã®æ¯èŒãè¡ãããããã£ã³ã°ååŒæã®ç¹åŸŽãæ確ã«ããŠããŸãã
çµ±èšçæ€å®ã®å®æœïŒã³ã«ã¢ãŽããã»ã¹ãã«ããæ€å®ãADFæ€å®ãKPSSæ€å®ãªã©ãè€æ°ã®çµ±èšçæ€å®ãè¡ãããšã§ãçµæã®çµ±èšçæææ§ã確èªããŠããŸãã
è€æ°ã®åæææ³ã®é©çšïŒãªã¿ãŒã³ååžã裟ã®éããæéäŸåæ§ãªã©ãç°ãªãåŽé¢ããåæãè¡ãããšã§ãçµæã®é å¥æ§ã確ä¿ããŠããŸãã
èŠèŠåïŒèªå·±çžé¢ããããããªã¿ãŒã³ååžã®ãã£ããã£ã³ã°ãªã©ãçµæãèŠèŠçã«è¡šçŸããããšã§ãçºèŠãããç¹åŸŽãæ確ã«ç€ºããŠããŸãã
ãããã®æ¹æ³ã«ãããç 究çµæã®ä¿¡é Œæ§ãšæå¹æ§ãé«ããŠããŸãããã ããèè ãã¯ç 究ã®éçïŒäŸïŒããŒã¿æéã®çãïŒãèªèããŠãããå°æ¥ã®ç 究ã®æ¹åæ§ã瀺ããŠããŸãã
è°è«ã¯ããïŒ
èè ãã¯ç 究çµæã«åºã¥ããŠä»¥äžã®ãããªè°è«ãå±éããŠããŸãïŒ
ãããã£ã³ã°ååŒæã®å¹çæ§ïŒéèåžå Žãšæ¯èŒããŠããããã£ã³ã°ååŒæãããå¹ççã§ããå¯èœæ§ã瀺åãããŠããŸããããã¯ããªã¿ãŒã³ååžã®ã軜ã裟ããèªå·±çžé¢ã®éãæžè¡°ãªã©ããæšæž¬ãããŠããŸãã
äºæž¬å¯èœæ§ã®å·®ç°ïŒã¹ããŒãã€ãã³ãã®çµæããäžççµæžã«åœ±é¿ãäžããåºæ¥äºãããæ¬è³ªçã«äºæž¬ããããå¯èœæ§ããããšè°è«ããŠããŸããããããéèåžå Žãšã¯ç°ãªããªã¿ãŒã³ååžã®ç¹æ§ã«ã€ãªãã£ãŠããå¯èœæ§ããããŸãã
åå è ã®ç¹æ§ïŒãããã£ã³ã°ååŒæã®åå è ããéèåžå Žã®åå è ããããªã¹ã¯åé¿åºŠãäœãå¯èœæ§ããããšææããŠããŸããããããããåçãªå©çãšæ倱ã®ååžã«ã€ãªãã£ãŠããå¯èœæ§ããããŸãã
ããŒã¿ã®å¶éïŒç 究ã§äœ¿çšããããŒã¿ã1ãæéã®ãã®ã§ãããé·æçãªåŸåãå£ç¯æ§ãæããããŠããªãå¯èœæ§ãããããšãèªèããŠããŸãã
ä»åŸã®ç 究æ¹åïŒããé·æã®ããŒã¿åæãç°ãªãã¹ããŒããåžå Žã®æ¯èŒããããã£ã³ã°ååŒæç¹æã®å£ç¯æ§ã®èª¿æ»ãªã©ãå°æ¥ã®ç 究課é¡ãææ¡ããŠããŸãã
ãããã®è°è«ã¯ããããã£ã³ã°ååŒæãšéèåžå Žã®éããç解ããä»åŸã®ç 究ã®æ¹åæ§ã瀺ãäžã§éèŠãªåœ¹å²ãæãããŠããŸãã
次ã«èªãã¹ãè«æã¯ïŒ
èè ãã¯çŽæ¥çã«æ¬¡ã«èªãã¹ãè«æãæšå¥šããŠããŸããããç 究ã®æèãšå°æ¥ã®æ¹åæ§ãã以äžã®è«æãé¢é£æ§ãé«ããšèããããŸãïŒ
Restocchi, V. (2018). "It Takes All Sorts: the Complexity of Prediction Markets." ãã®è«æã¯äºæž¬åžå Žã®ãªã¿ãŒã³ååžãåæããŠãããæ¬ç 究ãšæ¯èŒããããšã§ããããã£ã³ã°ååŒæã®ç¹æ§ãããæ·±ãç解ã§ããå¯èœæ§ããããŸãã
Smith, M. A., Paton, D., & Williams, L. V. (2006). "Market efficiency in person-to-person betting." Economica, 73(292), 673-689. ãã®è«æã¯ãããã£ã³ã°ååŒæã®å¹çæ§ãç 究ããŠãããæ¬ç 究ã®å¹çæ§ã«é¢ããè°è«ãè£å®ããå¯èœæ§ããããŸãã
Franck, E., Verbeek, E., & NÃŒesch, S. (2010). "Prediction accuracy of different market structures - bookmakers versus a betting exchange." International Journal of Forecasting, 26(3), 448-459. ãã®è«æã¯ãããã£ã³ã°ååŒæã®äºæž¬ç²ŸåºŠãåæããŠãããæ¬ç 究ã®å¹çæ§ã«é¢ããè°è«ãšé¢é£ä»ããããšãã§ããŸãã
Hardiman, S. J., Richmond, P., & Hutzler, S. (2010). "Long-range correlations in an online betting exchange for a football tournament." New Journal of Physics, 12(10), 105001. ãã®è«æã¯ãµãã«ãŒåžå Žã«ããããããã£ã³ã°ååŒæã®é·æçžé¢ãç 究ããŠãããæ¬ç 究ã®æéäŸåæ§åæãè£å®ããå¯èœæ§ããããŸãã
ãããã®è«æãèªãããšã§ããããã£ã³ã°ååŒæã®çµ±èšçç¹æ§ãå¹çæ§ã«ã€ããŠãããå æ¬çãªç解ãåŸãããšãã§ããã§ãããã
ãã®è«æããã銬åžè³Œå ¥è ã®è¡åã«ã€ããŠèªã¿åããããšã¯ãããªæãã ãšæãã
- ãªã¹ã¯ç®¡çïŒæ¥µç«¯ãªåã¡è² ããå°ãªããæ éã«ãããããåŸåãããã
- æ å ±æŽ»çšïŒå©çšå¯èœãªæ å ±ãå¹æçã«äœ¿ããåççãªå€æãããåŸåãããã
- äºæž¬èœåïŒã¹ããŒãã«é¢ããç¥èãåæåã掻ãããŠããããè¡ã£ãŠããå¯èœæ§ãé«ãã
- ãã©ã³ã¹ïŒé床ã«å©çãè¿œæ±ãããæ倱ãæããããããæ¯èŒçãã©ã³ã¹ã®åããæ 床ã§ãããããŠããã
- åŠç¿é©å¿åïŒçµéšããåŠã³ãåžå Žã®ååã«é©å¿ããèœåãããã
- ã³ãã¥ããã£æ§ïŒç¬èªã®ç¥èãçµéšãå ±æããç¹æ®ãªã³ãã¥ããã£ã圢æããŠããå¯èœæ§ãããã
- 嚯楜å¿åïŒå€ãã®åå è ãé床ã®ãªã¹ã¯ãé¿ããäž»ã«åš¯æ¥œãšããŠãããã£ã³ã°ã楜ããã§ããå¯èœæ§ãããã
ããªãã¥ãã¥ãšã«æ¹åŒã§ã¯éŠ¬åžè³Œå ¥è éã§ã®åãåããªã®ã§ãæ å ±ã®é察称æ§ã£ãŠã®ã¯ãªããºãçãèŠçŽ ãšããŠå€§ããã£ããšæãããã ããã®ããããä»ã¯åãŸã£ãŠããŠããŠãæ¥æ¬ã§ãæè¿ã¯ãªããºãæžããªã£ãŠããŠãããšæãããèæçã«ã¯çŽåŸã§ãããšãããããã
Optimal speed in Thoroughbred horse racing
以äžClaude-3.5-Sonnetã«ããèœåããã³ãã
ãµã©ãã¬ãã競走銬ã®ã¬ãŒã¹ã«ãããæé©é床
ã©ããªãã®ïŒ
ãã®è«æã¯ã競走銬ïŒãµã©ãã¬ããïŒãã¬ãŒã¹äžã«ã©ã®ããã«é床ã調æŽãã¹ããããæ°åŠçã¢ãã«ãçšããŠåæãããã®ã§ãã競走銬ã¯å šåçŸèµ°ãç¶æã§ããªããããããŒã¹é åãšæåŸã®ã¹ããŒãã®ã¿ã€ãã³ã°ãéèŠã§ããèè ãã¯ãååŠããšãã«ã®ãŒä»£è¬ïŒæé žçŽ ã»ç¡é žçŽ ïŒãéåå¶åŸ¡ãèæ ®ããã¢ãã«ãæ§ç¯ããŸããããã®ã¢ãã«ã¯ãé床ãæšé²åãç¡é žçŽ ãšãã«ã®ãŒã«é¢ããé£ç«åžžåŸ®åæ¹çšåŒãããªããæé©å¶åŸ¡åé¡ãšããŠè§£ãããŸãããã©ã³ã¹ã®ã·ã£ã³ãã£ãŒç«¶éŠ¬å Žã§ã®ã¬ãŒã¹ããŒã¿ãçšããŠãã©ã¡ãŒã¿ãåå®ããæ°å€ã·ãã¥ã¬ãŒã·ã§ã³ã«ãããã¬ãŒã¹äžã®æé©ãªé床ãé žçŽ æåéã®å€åããšãã«ã®ãŒæ¶è²»ãæšé²åãäºæž¬ããŠããŸããããã«ãã³ãŒã¹ã®å°åœ¢ïŒé«äœå·®ãã«ãŒãïŒã«å¿ããŠã銬ãã©ã®ããã«åªåãšé床ãå€ããã¹ãããäºæž¬ããŠããŸãã
å è¡ç 究ãšæ¯ã¹ãŠã©ãããããïŒ
ãã®ç 究ã®é©æ°çãªç¹ã¯ã競走銬ã®ã¬ãŒã¹æŠç¥ãæ°åŠçã«ã¢ãã«åããæé©åããç¹ã§ããå è¡ç 究ã§ã¯ãäž»ã«ãã¬ãããã«äžã§ã®æž¬å®ããã¬ãŒã¹äžã®åçŽãªé床åæã«çãŸã£ãŠããŸãããäŸãã°ãEaton et al. (1995)ãRose et al. (1988)ã¯ããã¬ãããã«äžã§ã®æå€§é žçŽ æåéïŒVO2maxïŒã枬å®ããŠããŸããããå®éã®ã¬ãŒã¹äžã®å®å šãªVO2ãããã¡ã€ã«ã¯äžæã§ãããSpence et al. (2012)ã¯ã¬ãŒã¹äžã®ããŒã¹æŠç¥ãåæããŠããŸããããšãã«ã®ãŒä»£è¬ãå°åœ¢ã®åœ±é¿ã¯èæ ®ããŠããŸãããæ¬ç 究ã¯ãå®éã®ã¬ãŒã¹ããŒã¿ãçšããŠãååŠããšãã«ã®ãŒä»£è¬ãéåå¶åŸ¡ãçµ±åããã¢ãã«ãæ§ç¯ããã¬ãŒã¹å šäœãéããŠã®æé©æŠç¥ãäºæž¬ããŠããç¹ã§ç»æçã§ããããã«ãåéãã«ãŒããã¬ãŒã¹æŠç¥ã«äžãã圱é¿ãå®éçã«åæããŠããç¹ãæ°ããè²¢ç®ã§ãã
æè¡ãææ³ã®èã¯ã©ãïŒ
ãã®ç 究ã®æ žå¿ã¯ã競走銬ã®ã¬ãŒã¹æŠç¥ãæé©å¶åŸ¡åé¡ãšããŠå®åŒåããç¹ã«ãããŸããå ·äœçã«ã¯ä»¥äžã®èŠçŽ ãéèŠã§ãïŒ
- é£ç«åžžåŸ®åæ¹çšåŒã¢ãã«ïŒé床ãæšé²åãç¡é žçŽ ãšãã«ã®ãŒã®æéå€åãèšè¿°
- ãšãã«ã®ãŒä»£è¬ã®èæ ®ïŒæé žçŽ ç³»ïŒVO2ïŒãšç¡é žçŽ ç³»ã®ãã©ã³ã¹ãã¢ãã«å
- å°åœ¢ã®åœ±é¿ïŒåéãã«ãŒããæšé²åãšé床ã«äžãã圱é¿ãçµã¿èŸŒã¿
- ãã©ã¡ãŒã¿åå®ïŒå®éã®ã¬ãŒã¹ããŒã¿ãçšããŠéŠ¬ã®ççåŠçãã©ã¡ãŒã¿ãæšå®
- æé©å¶åŸ¡åé¡ã®è§£æ³ïŒBOCOPããŒã«ããã¯ã¹ã䜿çš
- ç²å矀æé©åïŒãã©ã¡ãŒã¿åå®ã®ããã®ãã¥ãŒãªã¹ãã£ãã¯æé©åææ³
ãããã®èŠçŽ ãçµã¿åãããããšã§ãã¬ãŒã¹è·é¢ãã³ãŒã¹åœ¢ç¶ã銬ã®èœåã«å¿ããæé©ãªããŒã¹æŠç¥ãå°åºããŠããŸããç¹ã«ãVO2ã®å€åãšãšãã«ã®ãŒæ¶è²»ã®ã¢ãã«åããçŸå®çãªã¬ãŒã¹æŠç¥ã®äºæž¬ãå¯èœã«ããŠããç¹ãéèŠã§ãã
ã©ããã£ãŠæå¹ã ãšæ€èšŒããïŒ
èè ãã¯ããã©ã³ã¹ã®ã·ã£ã³ãã£ãŒç«¶éŠ¬å Žã§è¡ãããå®éã®ã¬ãŒã¹ããŒã¿ãçšããŠã¢ãã«ã®æ€èšŒãè¡ã£ãŠããŸããå ·äœçã«ã¯ã以äžã®æé ã§æ€èšŒãå®æœããŸããïŒ
- ããŒã¿éžæïŒä»ã®éŠ¬ãšã®çžäºäœçšãå°ãªããç¹å¥ãªæŠç¥ãåã£ãŠããªãã¬ãŒã¹ãéžæ
- ãã©ã¡ãŒã¿åå®ïŒéžæãããã¬ãŒã¹ããŒã¿ãçšããŠãã¢ãã«ã®ãã©ã¡ãŒã¿ãåå®
- ã·ãã¥ã¬ãŒã·ã§ã³ïŒåå®ããããã©ã¡ãŒã¿ãçšããŠãæé©ãªé床ãããã¡ã€ã«ãèšç®
- æ¯èŒæ€èšŒïŒã·ãã¥ã¬ãŒã·ã§ã³çµæãšå®éã®ã¬ãŒã¹ããŒã¿ïŒé床ãããã¡ã€ã«ïŒãæ¯èŒ
- è€æ°è·é¢ã§ã®æ€èšŒïŒ1300mã1900mã2100mã®ã¬ãŒã¹ã§ã¢ãã«ã®é©çšæ§ã確èª
- äºæž¬ã®åŠ¥åœæ§ïŒåŸãããVO2ãããã¡ã€ã«ãç¡é žçŽ ãšãã«ã®ãŒã®æ¶è²»ãã¿ãŒã³ããæ¢ç¥ã®ççåŠçç¥èŠãšæŽåçãã確èª
çµæãšããŠãã¢ãã«ã«ããäºæž¬ãå®éã®ã¬ãŒã¹ããŒã¿ãšããäžèŽãããŸãççåŠçã«ã劥åœãªçµæãåŸãããããšãããã¢ãã«ã®æå¹æ§ã瀺ãããŠããŸããç¹ã«ãã«ãŒããåéãã¬ãŒã¹æŠç¥ã«äžãã圱é¿ãå®éçã«äºæž¬ã§ããç¹ããã¢ãã«ã®åŒ·ã¿ãšããŠç€ºãããŠããŸãã
è°è«ã¯ããïŒ
è«æã§ã¯ä»¥äžã®ãããªè°è«ç¹ãæããããŠããŸãïŒ
- VO2ãããã¡ã€ã«ïŒã¢ãã«ã¯ãã¬ãŒã¹äžã®VO2ã®å€åãäºæž¬ããŠãããããã¯çŽæ¥æž¬å®ãå°é£ãªéèŠãªæ å ±ã§ããç¹ã«ãVO2maxã«å°éããæéããã¬ãŒã¹çµç€ã§ã®VO2ã®äœäžéå§ææã«ã€ããŠæ°ããªç¥èŠãæäŸããŠããŸãã
- ãšãã«ã®ãŒä»£è¬ïŒæé žçŽ ç³»ãšç¡é žçŽ ç³»ã®ãšãã«ã®ãŒäŸçµŠãã©ã³ã¹ã«ã€ããŠãã¬ãŒã¹è·é¢ã«ããéããå®éçã«ç€ºããŠããŸããããã¯ããã¬ãŒãã³ã°æŠç¥ã®æé©åã«è²¢ç®ããå¯èœæ§ããããŸãã
- æé©ããŒã¹æŠç¥ïŒã¢ãã«ã¯ã匷ãã¹ã¿ãŒãããã·ã¥ã®éèŠæ§ã瀺åããŠããŸããã銬ã®èœåã«å¿ããŠé©åã«èª¿æŽããå¿ èŠæ§ãææããŠããŸãã
- å°åœ¢ã®åœ±é¿ïŒåéãã«ãŒããåã«å±æçãªåœ±é¿ã§ã¯ãªããã¬ãŒã¹å šäœã®æŠç¥ã«å€§ãã圱é¿ããããšã瀺ããŠãããã³ãŒã¹ç¹æ§ã«å¿ããæŠç¥ã®éèŠæ§ã匷調ããŠããŸãã
- ã¢ãã«ã®éçïŒçŸæ®µéã§ã¯ä»éŠ¬ãšã®çžäºäœçšïŒãã©ããã£ã³ã°å¹æãªã©ïŒã銬ã®å¿ççèŠå ãèæ ®ããŠããªãç¹ãéçãšããŠæããããŠããŸãã
- å¿çšå¯èœæ§ïŒåã ã®éŠ¬ã®èœåè©äŸ¡ããé©æ§è·é¢ã®äºæž¬ãªã©ãå®è·µçãªå¿çšã®å¯èœæ§ã«ã€ããŠè°è«ããŠããŸãã
ãããã®è°è«ã¯ãã¢ãã«ã®æçšæ§ãšä»åŸã®ç 究æ¹åæ§ã瀺åããŠããŸãã
次ã«èªãã¹ãè«æã¯ïŒ
æ¬ç 究ã®çºå±ãé¢é£åéã®ç解ãæ·±ããããã«ã以äžã®è«æãåèã«ãªããšèããããŸãïŒ
- Spence AJ, et al. (2012) "Speed, pacing strategy and aerodynamic drafting in Thoroughbred horse racing" - 競走銬ã®ããŒã¹æŠç¥ãšãã©ããã£ã³ã°å¹æã«é¢ããç 究
- Hanon C, Thomas C. (2011) "Effects of optimal pacing strategies for 400-, 800-, and 1500-m races on the VO2 response" - ããã®äžè·é¢èµ°ã«ãããããŒã¹æŠç¥ãšVO2å¿çã®é¢ä¿ãåæ
- Self ZT, et al. (2012) "Speed and incline during Thoroughbred horse racing: racehorse speed supports a metabolic power constraint to incline running but not to decline running" - 競走銬ã®ã¹ããŒããšåéã®é¢ä¿ãåæ
- Aftalion A, Trélat E. (2020) "Pace and motor control optimization for a runner" - ã©ã³ããŒã®ããŒã¹ãšéåå¶åŸ¡ã®æé©åã«é¢ããçè«ç 究
- Cottin F, et al. (2010) "Oxygen consumption and gait variables of Arabian endurance horses measured during a field exercise test" - æä¹ èµ°éŠ¬ã®é žçŽ æ¶è²»éãšæ©æ§å€æ°ã®é¢ä¿ãåæ
- Sih A, et al. (2004) "Behavioral syndromes: an ecological and evolutionary overview" - åç©ã®è¡åç¹æ§ã«é¢ããçæ åŠçã»é²ååŠçç 究
ãããã®è«æã¯ã競走銬ã®ããã©ãŒãã³ã¹ããšãã«ã®ãŒä»£è¬ãæé©åã¢ãã«ãè¡åç¹æ§ãªã©ãæ¬ç 究ãšé¢é£ããæ§ã ãªåŽé¢ãã«ããŒããŠããŸãã
以äžã®æžç±ã«ã¬ãŒã¹ã«ãããããŒã¹é åããšãã«ã®ãŒæ¶è²»ãªã©ããŸãšããããŠããã®ã ãã©ãå 容çã«ã¯äžèŽããæãå€ãã®ã§çŽåŸã
ã¡ãã£ãšé¢çœãã®ã§è²ã èããŠã¿ãã
éšæã¯ã©ãèšãããšãèããŠä¹ãã°ããïŒå
·äœçãªãã®ãããã°æããŠã
éšæãèæ ®ãã¹ãå ·äœçãªç¹ã«ã€ããŠãè«æã®å 容ã«åºã¥ããŠä»¥äžã«ãŸãšããŸãïŒ
ã¹ã¿ãŒãããã·ã¥ã®éèŠæ§
- è«æã«ãããšã匷ãã¹ã¿ãŒãããã·ã¥ãéèŠã§ããããã«ããã銬ã®VO2ïŒé žçŽ æåéïŒãçŽ æ©ãæ倧å€ã«å°éãããããšãã§ããŸãã
- å ·äœçã«ã¯ãã¹ã¿ãŒãåŸ200-300ã¡ãŒãã«ã§æé«é床ã«éããããã«ããŒã¹ã調æŽããããšãæšå¥šãããŠããŸãã
銬ã®èœåã«å¿ããããŒã¹é å
- ãã ããã¹ã¿ãŒãããã·ã¥ã¯éŠ¬ã®èœåã«å¿ããŠé©åã«èª¿æŽããå¿ èŠããããŸãã
- 銬ã®ç¡é žçŽ æ§ãšãã«ã®ãŒäŸçµŠèœåïŒe0ãã©ã¡ãŒã¿ïŒãšVO2maxãèæ ®ããã¬ãŒã¹åŸåã§æ¥µç«¯ã«å€±éããªããããªããŒã¹é åãéèŠã§ãã
ã³ãŒã¹å°åœ¢ã®åœ±é¿ãèæ ®
- åéãã«ãŒããã¬ãŒã¹å šäœã®æŠç¥ã«å€§ãã圱é¿ããããšãèªèããå¿ èŠããããŸãã
- äžãåã§ã¯é床ãèœã¡ãããããã®ååŸã§ããŒã¹ã調æŽããŸãã
- ã«ãŒãã§ã¯é å¿åã®åœ±é¿ã§æšé²åãå¶éããããããã«ãŒãã«å ¥ãåã«ããé床ãèœãšããã«ãŒãæãã§å éããã®ãå¹æçã§ãã
ã¬ãŒã¹è·é¢ã«å¿ãããšãã«ã®ãŒé å
- çè·é¢ïŒäŸïŒ1300mïŒã§ã¯ç¡é žçŽ æ§ãšãã«ã®ãŒã®å²åãé«ããªããŸãïŒè«æã§ã¯çŽ47%ïŒã
- äžè·é¢ïŒäŸïŒ1900mã2100mïŒã§ã¯æé žçŽ æ§ãšãã«ã®ãŒã®å²åãå¢å ããŸãïŒè«æã§ã¯çŽ62-68%ïŒã
- ã¬ãŒã¹è·é¢ã«å¿ããŠãã¹ã¿ãŒãããã·ã¥ã®åŒ·ããšäžç€ã®ããŒã¹ç¶æã®ãã©ã³ã¹ã調æŽããŸãã
ã©ã¹ãã¹ããŒãã®ã¿ã€ãã³ã°
- è«æã«ãããšãæ®ãã®ç¡é žçŽ æ§ãšãã«ã®ãŒãåæå€ã®çŽ30%ã«ãªã£ãæç¹ã§VO2ãäœäžãå§ããŸãã
- ãããã©ã¹ãã¹ããŒããä»æããã¹ãæé©ãªã¿ã€ãã³ã°ãšãããŠããŸãã
- å ·äœçã«ã¯ããŽãŒã«å500-600ã¡ãŒãã«ä»è¿ã§ã®ã¹ããŒããå¹æçã§ããå¯èœæ§ãé«ãã§ãã
ä»éŠ¬ãšã®äœçœ®é¢ä¿
- è«æã§ã¯çŽæ¥æ±ãããŠããŸãããããã©ããã£ã³ã°ïŒé¢šããïŒå¹æãèæ ®ããããšãéèŠã§ãã
- ç¹ã«äžé·è·é¢ã¬ãŒã¹ã§ã¯ãä»éŠ¬ã®åŸãã«ã€ããŠãšãã«ã®ãŒãæž©åããæåŸã®çŽç·ã§æãåºãæŠç¥ãå¹æçãããããŸããã
銬ã®å¿çç¶æ ãžã®é æ ®
- ãããè«æã§ã¯çŽæ¥æ±ãããŠããŸãããã銬ã®æ§æ Œãå¿çç¶æ ã«å¿ããã©ã€ãã£ã³ã°ãéèŠã§ãã
- äŸãã°ãç¥çµè³ªãªéŠ¬ã«ã¯èœã¡çããããŒã¹é åã§ãéäºå¿ã®åŒ·ã銬ã«ã¯ä»éŠ¬ãšã®ç«¶ãåããæèããã©ã€ãã£ã³ã°ãå¿ããããªã©ã®å·¥å€«ãå¿ èŠã§ãããã
ãããã®ç¹ãç·åçã«èæ ®ããªãããåã ã®éŠ¬ã®ç¹æ§ãšã¬ãŒã¹ã®ç¶æ³ã«å¿ããŠæé©ãªæŠç¥ãç«ãŠãããšããéšæã«ã¯æ±ããããŸãã
ã³ãŒã¹å°åœ¢ã®åœ±é¿ãèæ
®ã£ãŠãšããããã£ãšè©³ãã
ã³ãŒã¹å°åœ¢ã®åœ±é¿ã«ã€ããŠã®è©³çŽ°ãªèå¯ã¯ããã®è«æã®éèŠãªè²¢ç®ã®äžã€ã§ããéšæãã³ãŒã¹å°åœ¢ãèæ ®ããéã®å ·äœçãªãã€ã³ãããè«æã®å 容ã«åºã¥ããŠããã«è©³ãã説æããŸãïŒ
åéã®åœ±é¿
- äžãåïŒ
- äžãåã§ã¯éŠ¬ã®é床ãèªç¶ã«äœäžããŸããè«æã®ã·ãã¥ã¬ãŒã·ã§ã³ã§ã¯ã3%ã®äžãåã§çŽ1-2 m/s ã®é床äœäžãèŠãããŸããã
- äžãåã®äœçœ®ã«ãã£ãŠæŠç¥ãå€ãããŸãïŒ
- ã¬ãŒã¹åºç€ã®äžãåïŒæ倧é床ã«å°éããã®ã劚ãããããéåžžããã匷ãã®ã¹ã¿ãŒãããã·ã¥ãå¿ èŠãããããŸããã
- ã¬ãŒã¹äžç€ã®äžãåïŒå¹³åé床ãäžããããããã®ååŸã§ããŒã¹ãäžããŠè£å®ããããšãå¹æçã§ãã
- ã¬ãŒã¹çµç€ã®äžãåïŒéŠ¬ããŸã ååãªãšãã«ã®ãŒãæã£ãŠããã°ãä»ã®éŠ¬ããã匷ãæåŸã®è¿œã蟌ã¿ãã§ããå¯èœæ§ããããŸãã
- äžãåïŒ
- äžãåã§ã¯èªç¶ã«é床ãäžãããŸããããããå©çšããŠãšãã«ã®ãŒãæž©åããããšãéèŠã§ãã
- ç¹ã«ã¬ãŒã¹åºç€ãäžç€ã®äžãåã§ã¯ãé床ã«é床ãäžããããªããã泚æãå¿ èŠã§ãã
ã«ãŒãã®åœ±é¿
- ã«ãŒãã§ã¯é å¿åã®åœ±é¿ã§æå¹ãªæšé²åãæžå°ããŸããè«æã®ã¢ãã«ã§ã¯ãããã f_tot = â(f^2 + c^2v^4) †f_M ãšããå¶çŽã§è¡šçŸããŠããŸãïŒf ã¯æšé²åãc ã¯ã«ãŒãã®æ²çãv ã¯é床ãf_M ã¯æ倧æšé²åïŒã
- ã«ãŒãã®äœçœ®ã«ãã圱é¿ïŒ
- ã¬ãŒã¹åºç€ã®ã«ãŒãïŒæé«é床ãžã®å°éãé ããããããã¹ã¿ãŒãããã·ã¥åŸã®å éãæ§ããã«ããå¿ èŠããããŸãã
- ã¬ãŒã¹äžç€ã®ã«ãŒãïŒå šäœçãªå¹³åé床ãžã®åœ±é¿ãæå°éã«ãªããããããã§ã®ããŒã¹èª¿æŽãéèŠã§ãã
- ã¬ãŒã¹çµç€ã®ã«ãŒãïŒæåŸã®ã¹ããŒãã®ã¿ã€ãã³ã°ã«åœ±é¿ãããããã«ãŒãåºå£ã§ã®å éãåè² ã®éµãšãªãå¯èœæ§ããããŸãã
- å ·äœçãªæŠç¥ïŒ
- ã«ãŒãã«å ¥ãåã«ããé床ãèœãšããã«ãŒãäžã¯äžå®ã®ããŒã¹ãä¿ã¡ãã«ãŒããåºããšããã§å éããã®ãå¹æçã§ãã
- è«æã®ã·ãã¥ã¬ãŒã·ã§ã³ã§ã¯ãååŸ100mã®ã«ãŒãã§çŽ1-1.5 m/s ã®é床äœäžãèŠãããŸããã
ã¯ããœã€ãïŒç·©åæ²ç·ïŒã®å©çš
- è«æã§ã¯ãçŽç·ãšã«ãŒããã€ãªãã¯ããœã€ãïŒæ²çãè·é¢ã«æ¯äŸããŠå¢å ããæ²ç·ïŒã®éèŠæ§ãææãããŠããŸãã
- ã¯ããœã€ãã§ã¯ã銬ãæ»ããã«é床ãå€åãããããšãã§ããŸãã
- éšæã¯ãã¯ããœã€ãåºéã§åŸã ã«äœå¢ãæŽããã«ãŒãã«åããããã«ãŒãåºå£ã§å éã®æºåããããããããšãã§ããŸãã
è€åçãªå°åœ¢å€åãžã®å¯Ÿå¿
- å®éã®ã³ãŒã¹ã§ã¯ãåéãšã«ãŒããçµã¿åããã£ãŠããããšãå€ãã§ãã
- äŸãã°ãè«æã§åæãããã·ã£ã³ãã£ãŒç«¶éŠ¬å Žã®2100mã³ãŒã¹ã§ã¯ãã¹ã¿ãŒãåŸããã«äžãåã®ã«ãŒããããããã®åŸäžãåã«ãªã£ãŠããŸãã
- ãã®ãããªè€åçãªå°åœ¢å€åã«å¯ŸããŠã¯ãåã ã®èŠçŽ ã®åœ±é¿ãç·åçã«èæ ®ããããŒã¹é åãå¿ èŠã§ãã
ã³ãŒã¹å šäœãéããŠã®æŠç¥
- è«æã¯ãå±æçãªå°åœ¢å€åãã¬ãŒã¹å šäœã®æŠç¥ã«åœ±é¿ãäžããããšã匷調ããŠããŸãã
- äŸãã°ãåºç€ã®äžãåã匷ãæ»ãããšãåŸåã®ããŒã¹ãç¶æã§ããªããªãå¯èœæ§ããããŸãã
- ãããã£ãŠãã³ãŒã¹å šäœã®å°åœ¢ãææ¡ãã銬ã®ãšãã«ã®ãŒæ¶è²»ãæé©åããç·åçãªæŠç¥ãéèŠã§ãã
ãããã®ç¹ãèæ ®ããããšã§ãéšæã¯éŠ¬ã®èœåãæ倧éã«åŒãåºããã³ãŒã¹ã®ç¹æ§ã掻ãããã¬ãŒã¹éã³ãããããšãã§ããŸãã
銬åžè³Œå
¥è
ã®ç«å Žã§èãããšããããã®å
容ã銬åžè³Œå
¥æã«æŽ»ããããšã¯ã§ããŸãããïŒ
銬åžè³Œå ¥è ã®ç«å Žããããã®ç 究ã®å 容ã掻çšããæ¹æ³ã«ã€ããŠèå¯ããŸãïŒ
ã³ãŒã¹é©æ§ã®è©äŸ¡
- å銬ã®éå»ã®ã¬ãŒã¹çµæããã³ãŒã¹åœ¢ç¶ïŒåãã«ãŒãã®äœçœ®ãªã©ïŒãšç §ããåãããŠåæããŸãã
- äŸãã°ãçµç€ã«æ¥ãªäžãåãããã³ãŒã¹ã§å¥œæ瞟ãæ®ããŠãã銬ã¯ã匷ãæä¹ åãç¡é žçŽ æ§èœåãæã£ãŠããå¯èœæ§ãé«ãã§ãã
è·é¢é©æ§ã®å€æ
- è«æã«ãããšãçè·é¢ïŒ1300mïŒã§ã¯ç¡é žçŽ æ§ãšãã«ã®ãŒã®è²¢ç®ãçŽ47%ãäžè·é¢ïŒ1900m-2100mïŒã§ã¯çŽ32-38%ã§ãã
- å銬ã®éå»ã®ããã©ãŒãã³ã¹ãããç¡é žçŽ æ§èœåãšæé žçŽ æ§èœåã®ãã©ã³ã¹ãæšæž¬ããé©æ£è·é¢ãå€æã§ããŸãã
ããŒã¹äºæ³
- ã³ãŒã¹ã®ç¹åŸŽïŒearlyéºããåãã¿ã€ããªã«ãŒããªã©ïŒãããã¬ãŒã¹å šäœã®ããŒã¹å±éãäºæ³ã§ããŸãã
- äŸãã°ãåºç€ã«äžãåãããå ŽåãããŒã¹ãæããããå¯èœæ§ãé«ããåŸç¶éŠ¬ã«æå©ãããããŸããã
éšæã®æŠè¡è©äŸ¡
- åéšæã®ã³ãŒã¹æ»ç¥æ³ãéå»ã®ã¬ãŒã¹ããåæãããã®ç 究ã®ç¥èŠãšç §ããåãããŠè©äŸ¡ããŸãã
- äŸãã°ãã«ãŒãååŸã§ããŸãããŒã¹ã³ã³ãããŒã«ã§ããéšæã¯é«è©äŸ¡ã§ããŸãã
ã³ã³ãã£ã·ã§ã³å€åã®åœ±é¿äºæž¬
- éŠ¬å Žç¶æ ã®å€åïŒè¯âéãªã©ïŒããå°åœ¢ã®åœ±é¿ãã©ãå€ããããäºæž¬ããŸãã
- äŸãã°ãééŠ¬å Žã§ã¯ã«ãŒãã®åœ±é¿ããã倧ãããªãå¯èœæ§ããããŸãã
å±éäºæ³
- ã¹ã¿ãŒãçŽåŸã®å°åœ¢ãããå銬ã®ããžã·ã§ã³åããäºæ³ã§ããŸãã
- äŸãã°ãã¹ã¿ãŒãåŸããã«æ¥ã«ãŒããããå Žåãå æ æå©ã®å±éãäºæ³ãããŸãã
æ«èã®äºæž¬
- ã³ãŒã¹åœ¢ç¶ãšè·é¢ãããã©ã¹ãã¹ããŒãã®ã¿ã€ãã³ã°ãšåŒ·ããäºæž¬ã§ããŸãã
- äŸãã°ããŽãŒã«å600mã§VO2ãäœäžãå§ãããšã®ç¥èŠã掻çšããããããã®è¿œã蟌ã¿åãè©äŸ¡ããŸãã
調æè©äŸ¡ã®ç²Ÿç·»å
- 調æå 容ããå®éã®ã¬ãŒã¹ã³ãŒã¹ã®å°åœ¢ç¹æ§ãšç §ããåãããŠè©äŸ¡ããŸãã
- äŸãã°ãæ¥åã®ãã調æã³ãŒã¹ã§ãã£ããä¹ã蟌ãŸããŠãã銬ã¯ãåæ§ã®ç¹åŸŽãæã€ç«¶éŠ¬å Žã§å¥œèµ°ããå¯èœæ§ãé«ããããããŸããã
é·æçãªéŠ¬ã®æé·è©äŸ¡
- 幎霢ã調æã«ãã銬ã®ççåŠçãã©ã¡ãŒã¿ïŒVO2maxãç¡é žçŽ æ§èœåãªã©ïŒã®å€åãæšæž¬ããå°æ¥æ§ãè©äŸ¡ããŸãã
穎銬ã®çºèŠ
- ã³ãŒã¹åœ¢ç¶ãšéŠ¬ã®ç¹æ§ïŒã¹ã¿ãŒãããã·ã¥ã®åŒ·ããæä¹ åãªã©ïŒã®ãããã³ã°ããããªããºã»ã©äººæ°ã§ãªã奜走銬ãèŠã€ããããå¯èœæ§ããããŸãã
ãããã®èŠç¹ãçµã¿åãããããšã§ããã粟床ã®é«ã銬åžè³Œå ¥ã®å€æææãšããããšãã§ããã§ãããããã ããå®éã®ã¬ãŒã¹ã§ã¯æ§ã ãªèŠå ãè€éã«çµ¡ã¿åãããããããã®ç¥èŠã¯åèçšåºŠã«çããç·åçãªå€æãå¿ èŠã§ãã
What AI can do for horse-racing ?
競銬ã«ãããŠAIãã§ããããšãšã¯ïŒ
1. ã©ããªãã®ïŒ
ãã®è«æã¯ã人工ç¥èœ(AI)ã競銬ã«ã©ã®ãããªåœ±é¿ãäžãããã«ã€ããŠè«ããŠããŸããèè ã¯ãã³ã³ãã¥ãŒã¿ããžã§ã³ãçµ±èšåŠç¿ãã²ãŒã çè«ãšãã3ã€ã®äž»èŠãªAIåéã競銬ã«å¿çšãããå¯èœæ§ãæ¢ã£ãŠããŸããå ·äœçã«ã¯ãã¬ãŒã¹çµæã®äºæž¬ç²ŸåºŠåäžã銬ãšéšæã®çžæ§è©äŸ¡ã銬ã®ææ ãè¡åã®ç解ãªã©ãæããããŠããŸãããŸãã競銬ãAIç 究ã®å®éšå ŽãšããŠæŽ»çšããããšã§ã人éäžå¿ã§ã¯ãªãåç©è¡åã®ç解ã«ã€ãªããå¯èœæ§ã瀺åããŠããŸããèè ã¯ããããã®æè¡ãä»åŸ10幎éã§ç«¶éŠ¬æ¥çã«å€§ããªå€é©ããããããšäºæž¬ããŠããŸãã
2. å è¡ç 究ãšæ¯ã¹ãŠã©ãããããïŒ
ãã®è«æã®é©æ°çãªç¹ã¯ã競銬ã«ãããAIå¿çšãå æ¬çã«è«ããŠããç¹ã§ããåŸæ¥ã®ç 究ã¯ãåã«ã¬ãŒã¹çµæã®äºæž¬ã«çŠç¹ãåœãŠãŠãããã®ãå€ãã£ãã®ã«å¯Ÿãããã®è«æã¯ç«¶éŠ¬å šäœãAIç 究ã®å®éšå ŽãšããŠæããèŠç¹ãæ瀺ããŠããŸããç¹ã«ãã³ã³ãã¥ãŒã¿ããžã§ã³ãçµ±èšåŠç¿ãã²ãŒã çè«ã®3åéã®èåãææ¡ããŠããç¹ãç¬ç¹ã§ãããŸãã競銬ããŒã¿ã®ç§å¿æ§ãé«ããšããæ¥çç¹æã®åé¡ãææãããªãŒãã³ãªç 究ã³ãã¥ããã£ã®å¿ èŠæ§ã蚎ããŠããç¹ã泚ç®ã«å€ããŸããããã«ãAIãéããŠéŠ¬ã®ææ ãè¡åãããæ·±ãç解ã§ããå¯èœæ§ã瀺åããŠããç¹ããåŸæ¥ã®ç 究ã«ã¯ãªãæ°ããèŠç¹ã§ãã
3. æè¡ãææ³ã®èã¯ã©ãïŒ
ãã®è«æã§ææ¡ãããŠããäž»èŠãªæè¡çã¢ãããŒãã¯ã3ã€ã®AIåéïŒã³ã³ãã¥ãŒã¿ããžã§ã³ãçµ±èšåŠç¿ãã²ãŒã çè«ïŒã®èåã§ããå ·äœçã«ã¯ã以äžã®3ã€ã®ããã³ãã£ã¢ãææ¡ããŠããŸãïŒ
- çµ±èšåŠç¿ãšã²ãŒã çè«ã®èåïŒåäºå®åæãçšããŠã銬ã®äºæž¬è¡åãšå®éã®è¡åãæ¯èŒããéšæã®æŠç¥çå€æãæœåºããã
- çµ±èšåŠç¿ãšã³ã³ãã¥ãŒã¿ããžã§ã³ã®èåïŒã¬ãŒã¹æ åãã銬ã®ææ ãåããåæããçµ±èšã¢ãã«ã«çµã¿èŸŒãã
- ã²ãŒã çè«ãšã³ã³ãã¥ãŒã¿ããžã§ã³ã®èåïŒãŠã©ãŒãã³ã°ã¢ããã銬äœã®å€èŠ³ãã銬䞻ã調æåž«ã®æåŸ ãæšæž¬ããã
ãããã®ææ³ã«ãããã¬ãŒã¹çµæã®äºæž¬ç²ŸåºŠåäžã ãã§ãªãã銬ãšéšæã®çžæ§è©äŸ¡ã銬ã®è¡åç解ãªã©ãããåºç¯ãªå¿çšãå¯èœã«ãªããšèè ã¯äž»åŒµããŠããŸãã4. ã©ããã£ãŠæå¹ã ãšæ€èšŒããïŒ
ãã®è«æã¯äž»ã«çè«çãªææ¡ãè¡ã£ãŠãããå ·äœçãªå®éšã«ããæ€èšŒã¯è¡ãããŠããŸãããããããèè ã¯æ¢åã®ç 究ãäºäŸãåŒçšããªãããææ¡ææ³ã®æœåšçãªæå¹æ§ã瀺åããŠããŸããäŸãã°ãåäºå®åæãçšããæŠç¥çå€æã®æœåºã«ã€ããŠã¯ããµãã«ãŒã§ã®é¡äŒŒç 究ãåŒçšããŠããŸãããŸããã³ã³ãã¥ãŒã¿ããžã§ã³ãçšãã銬ã®ææ åæã«ã€ããŠã¯ã銬ã®è¡šæ ãè³ã®äœçœ®ãã泚æåãæšæž¬ã§ãããšããå è¡ç 究ãåŒçšããŠããŸãã銬ãšéšæã®çžæ§è©äŸ¡ã«ã€ããŠã¯ã16é ã®éŠ¬ãš21人ã®éšæã察象ãšããå°èŠæš¡ãªç 究ãäŸã«æããAIãçšããããšã§å€§èŠæš¡ãªããŒã¿ã»ããã§ã®åæãå¯èœã«ãªããšäž»åŒµããŠããŸãããããããããã®ææ¡ã®å®éã®æå¹æ§ãæ€èšŒããã«ã¯ãä»åŸã®å®èšŒç 究ãå¿ èŠã§ããããšãèè ã¯èªããŠããŸãã
5. è°è«ã¯ããïŒ
èè ã¯ãAIã競銬ã«äžãã圱é¿ã«ã€ããŠãããã€ãã®éèŠãªè°è«ç¹ãæèµ·ããŠããŸãããŸããAIã«ããäºæž¬ç²ŸåºŠã®åäžãããããã£ã³ã°åžå Žã«ã©ã®ãããªåœ±é¿ãäžããããšããç¹ã§ããèè ã¯ãããæ£ç¢ºãªãªããºèšå®ãå¯èœã«ãªãäžæ¹ã§ãäžéšã®AIæè¡ãæã€éå£ãåžå Žãæ¯é ããå¯èœæ§ãææããŠããŸãã次ã«ãAIãçšãã銬ã®ææ ãè¡åã®ç解ããåç©çŠç¥ã®åäžã«ã€ãªããå¯èœæ§ãè«ããŠããŸããäŸãã°ãéã®äœ¿çšã銬ã®ã¹ããŒãã«åœ±é¿ãäžããªããšããç 究çµæãåŒçšããAIãããããç¥èŠãããã«æ·±ããå¯èœæ§ã瀺åããŠããŸãããŸããAIã«ãã競銬åæãã人éäžå¿ã§ã¯ãªãåç©è¡åã®ç解ã«ã€ãªããå¯èœæ§ãè°è«ãããŠããŸããæåŸã«ãããããAIæè¡ã®çºå±ãå ¬éãããç 究ã³ãã¥ããã£ã§è¡ãããã¹ããããããšãç§çãªå©çã®ããã«ç§å¯è£ã«è¡ãããã¹ãããšããå«ççãªåé¡ãæèµ·ãããŠããŸãã
6. 次ã«èªãã¹ãè«æã¯ïŒ
ãã®è«æã®å 容ãããã«æ·±ãç解ããé¢é£ããç 究ãæ¢ãããã«ã¯ã以äžã®è«æãèªãããšããå§ãããŸãïŒ
- Tuyls, K., et al. (2021). "Game plan: What AI can do for football, and what football can do for AI." ãã®è«æã¯ããµãã«ãŒã«ãããAIå¿çšãè«ããŠãããæ¬è«æã®å€ãã®ã¢ã€ãã¢ã®åºç€ãšãªã£ãŠããŸãã
- Benter, B. (1994). "Computer Based Horse Race Handicapping and Wagering Systems: A Report." 競銬ã«ãããã³ã³ãã¥ãŒã¿ãŒäºæž¬ã·ã¹ãã ã®å€å žçãªè«æã§ãã
- McGreevy, P. D., et al. (2012). "Whip Use by Jockeys in a Sample of Australian Thoroughbred RacesâAn Observational Study." éã®äœ¿çšãšéŠ¬ã®ããã©ãŒãã³ã¹ã®é¢ä¿ãç 究ããè«æã§ãAIãçšããåç©è¡åç解ã®å¯èœæ§ãæ¢ãäžã§åèã«ãªããŸãã
- Wathan, J., & McComb, K. (2014). "The eyes and ears are visual indicators of attention in domestic horses." 銬ã®ææ ã泚æåãèŠèŠçã«åæããç 究ã§ãã³ã³ãã¥ãŒã¿ããžã§ã³ã®å¿çšå¯èœæ§ã瀺åããŠããŸãã
çµ±èšåŠç¿ã«ã€ããŠã¯ããã¯ãäžè¬çã«ãªã£ãŠããŠããªããªããªããºãžã®åœ±é¿ãåºã€ã€ããæãã³ã³ãã¥ãŒã¿ããžã§ã³ã¯é¢çœãããªãã ãã©ãåŠç¿ããŒã¿ããããã
ã²ãŒã çè«ãã€ãã€ãããããªãã£ãã®ã§èããŠã¿ãã
ãã®æèã§ã®ã²ãŒã çè«ã¯ã©ããããã®ã§ãããïŒ
ãã®æèã§ã®ã²ãŒã çè«ã¯ãäž»ã«ç«¶éŠ¬ã«ãããæŠç¥çãªææ決å®ãåæã»ã¢ãã«åããããã«é©çšãããŠããŸãã以äžã«ãè«æã§èšåãããŠããã²ãŒã çè«ã®äž»ãªé©çšé åããŸãšããŸãïŒ
- ã·ãŒãºã³ã¬ãã«ã®æŠç¥ïŒ
- 銬䞻ã調æåž«ããã©ã®ã¬ãŒã¹ã«åºèµ°ãããã決å®ããéã®æŠç¥åæã
- åºèµ°ã«ããã¡ãªããïŒè³éç²åŸã銬ã®äŸ¡å€åäžïŒãšãã¡ãªããïŒæªæã®ãªã¹ã¯ãã³ã¹ãïŒã®ãã©ã³ã¹ãèæ ®ã
- ã¬ãŒã¹ã¬ãã«ã®æŠç¥ïŒ
- ã¬ãŒã¹å šäœã®æŠç¥ç«æ¡ïŒäŸïŒããŒã¹é åãããžã·ã§ãã³ã°ïŒã
- çè·é¢èµ°ãšé·è·é¢èµ°ã§ã®æé©æŠç¥ã®éããåæã
- ã¢ã¯ã·ã§ã³ã¬ãã«ã®æŠç¥ïŒ
- ãã©ããã£ã³ã°ïŒä»ã®éŠ¬ã®åŸãã«ä»ããŠé¢šã®æµæãæžããïŒãšã¢ã¿ãã¯ïŒå é ã«ç«ã€ïŒã®ã¿ã€ãã³ã°æ±ºå®ã
- éã®äœ¿çšã«é¢ããæŠç¥çå€æã
- éšæãšéŠ¬ã®çžäºäœçšïŒ
- éšæã®æŠç¥çå€æãšéŠ¬ã®çççèœåã®ãã©ã³ã¹ãåæã
- 人éïŒéšæïŒãšåç©ïŒéŠ¬ïŒã®ææ決å®ã®çžäºäœçšãã¢ãã«åã
- è€æ°ãšãŒãžã§ã³ãã®çžäºäœçšïŒ
- ã¬ãŒã¹äžã®å銬ã»éšæã®ãã¢ãç¬ç«ãããšãŒãžã§ã³ããšããŠæ±ãããã®çžäºäœçšãåæã
- ä»ã®ç«¶äºçžæã®è¡åãäºæž¬ããæé©ãªæŠç¥ãéžæããããã»ã¹ãã¢ãã«åã
- åäºå®åæïŒ
- äºæž¬ããã銬ã®è¡åãšå®éã®è¡åãæ¯èŒããéšæã®æŠç¥çå€æãæœåºã
- ãã®åæã«ããã銬ã®çççããã©ãŒãã³ã¹ãšéšæã®æŠç¥çå€æãåé¢ããŠè©äŸ¡ããããšãå¯èœã«ã
ãããã®é©çšã«ãããã²ãŒã çè«ã¯ç«¶éŠ¬ã«ãããè€éãªææ決å®ããã»ã¹ãç解ããã¢ãã«åããããã®åŒ·åãªããŒã«ãšãªã£ãŠããŸããããã«ãAIæè¡ãšçµã¿åãããããšã§ãããæŽç·ŽãããæŠç¥åæãäºæž¬ãå¯èœã«ãªããšè«æã¯ç€ºåããŠããŸãã
ãªãã»ã©ãã銬åžè³Œå ¥ã®ç«å Žã ãšã©ããïŒ
銬åžè³Œå
¥è
ã®è³Œå
¥å€æã«å¿çšããããšã¯ã§ããŸããïŒ
ã¯ããã²ãŒã çè«ãšAIã®çµã¿åããã¯ã銬åžè³Œå ¥è ã®è³Œå ¥å€æã«éåžžã«æå¹ã«å¿çšã§ããå¯èœæ§ããããŸãã以äžã«ããã®å¿çšå¯èœæ§ãããã€ãæããŠã¿ãŸãããïŒ
ãªããºåæïŒ
ã²ãŒã çè«ãçšããŠãä»ã®è³Œå ¥è ã®è¡åãäºæž¬ãããªããºã®å€åãåæããããšãã§ããŸããããã«ãããããªãŒããŒã¬ã€ãïŒå®éã®åçã«æ¯ã¹ãŠãªããºãé«ããã銬ïŒãèŠã€ãåºãå¯èœæ§ãé«ãŸããŸããæ å ±ã®é察称æ§ã®å©çšïŒ
ã²ãŒã çè«ã¯ãæ å ±ã®é察称æ§ãããç¶æ³ã§ã®ææ決å®ãåæããã®ã«é©ããŠããŸãã銬åžè³Œå ¥è ãæã€ç¬èªã®æ å ±ïŒäŸïŒéŠ¬ã®èª¿åã«é¢ããç®èŠæ å ±ïŒããã©ã®ããã«è³ãã«åæ ãããã¹ããã®å€æã«åœ¹ç«ã¡ãŸããè€åçãªè³ãæŠç¥ã®æé©åïŒ
ååãè€åã銬é£ãªã©æ§ã ãªè³ãæ¹ãããäžã§ãæåŸ å€ãæ倧åããçµã¿åãããã²ãŒã çè«çã¢ãããŒãã§åæã§ããŸããä»ã®è³Œå ¥è ã®è¡åäºæž¬ïŒ
倧å£ã®è³ããå ¥ã£ãéã®åžå Žã®åå¿ãªã©ãä»ã®è³Œå ¥è ã®è¡åãã¢ãã«åãäºæž¬ããããšã§ãããæŠç¥çãªè³Œå ¥å€æãå¯èœã«ãªããŸãããªã¹ã¯ç®¡çïŒ
ã²ãŒã çè«çã¢ãããŒãã«ãããé·æçãªåçãæ倧åãã€ã€ãªã¹ã¯ã管çããæŠç¥ãç«ãŠãããšãã§ããŸããããªãã¥ãŒãã¥ãšã«æ¹åŒã®åæïŒ
競銬ã®å€ãã§æ¡çšãããŠããããªãã¥ãŒãã¥ãšã«æ¹åŒïŒè³ãæå士ã§åã¡è² ãã競ãæ¹åŒïŒã¯ãã²ãŒã çè«çåæã«é©ããŠããŸããä»ã®è³Œå ¥è ã®è¡åãèªåã®ææ»éã«çŽæ¥åœ±é¿ããããããã®çžäºäœçšãã¢ãã«åããããšã§æå©ãªè³ããå¯èœã«ãªãå¯èœæ§ããããŸããAIãšã®çµ±åïŒ
è«æã§ææ¡ãããŠããããã«ãã²ãŒã çè«ãšAIïŒç¹ã«æ©æ¢°åŠç¿ãã³ã³ãã¥ãŒã¿ããžã§ã³ïŒãçµã¿åãããããšã§ãããé«åºŠãªåæãå¯èœã«ãªããŸããäŸãã°ïŒ
- ã¬ãŒã¹æ åãããªã¢ã«ã¿ã€ã ã§éŠ¬ã®ç¶æ ãåæ
- éå»ã®ããŒã¿ããåéšæã»éŠ¬ã®ç¹åŸŽãåŠç¿ããçŸåšã®ã¬ãŒã¹æ¡ä»¶ã§ã®æé©æŠç¥ãäºæž¬
- SNSãªã©ã®éæ§é åããŒã¿ãã銬ã®äººæ°åºŠãåæãããªããºã®å€åãäºæž¬
- å«ççèæ ®ïŒ
ãã ããããããé«åºŠãªåææè¡ã®äœ¿çšã«ã¯å«ççãªåé¡ã䌎ããŸããè«æã§ãææãããŠããããã«ãäžéšã®æè¡ãæã€éå£ãåžå Žãæ¯é ããå¯èœæ§ããããŸãããŸããé床ã®ã®ã£ã³ãã«äŸåãå©é·ããå¯èœæ§ãããããã責任ãã䜿çšãæ±ããããŸãããããã®å¿çšã«ããã銬åžè³Œå ¥è ã¯ããæ å ±ã«åºã¥ããæŠç¥çãªå€æãå¯èœã«ãªããšèããããŸãããã ãã競銬ã®ãããªè€éãªã·ã¹ãã ã§ã¯åžžã«äºæž¬äžå¯èœãªèŠçŽ ãååšããããããããã®åæã¯ãããŸã§ãå€æã®äžå©ãšããŠæããã¹ãã§ãããã
LLMã掻ãããªããïŒ
ã§ã¯ã倧èŠæš¡èšèªçæã¢ãã«ã䜿ããšãããšãçµ±èšåŠç¿ã»ã³ã³ãã¥ãŒã¿ããžã§ã³ã»ã²ãŒã çè«ããããã§ã©ããã掻çšãèããããã§ããããïŒ
倧èŠæš¡èšèªçæã¢ãã«ïŒLLMïŒã®ç«¶éŠ¬åæã«ããã掻çš:
- çµ±èšåŠç¿ã§ã®æŽ»çš:
- ããŒã¿çµ±åãšååŠç:
LLMã䜿ã£ãŠãæ§ã ãªåœ¢åŒãèšèªã§èšè¿°ããã銬ã®ããŒã¿ïŒéå»ã®æ瞟ãè¡çµ±æ å ±ã調æå ±åãªã©ïŒãçµ±åããæ§é åãããããŒã¿ã«å€æã§ããŸããããã«ãããç°ãªããœãŒã¹ããã®ããŒã¿ãäžè²«ãã圢åŒã§åæã«å©çšã§ããããã«ãªããŸãã- ç¹åŸŽæœåº:
LLMã¯èªç¶èšèªã§æžãããæ å ±ïŒäŸïŒèª¿æåž«ã®ã³ã¡ã³ããã¬ãŒã¹åæèšäºïŒããéèŠãªç¹åŸŽãæœåºããæ°å€ããŒã¿ãšçµã¿åãããŠåæã«æŽ»çšã§ããŸãã- ã¢ãã«è§£é:
è€éãªçµ±èšã¢ãã«ã®çµæãèªç¶èšèªã§èª¬æããããšã§ãå°é家ã§ãªã人ã«ããããããã圢ã§åæçµæãæ瀺ã§ããŸãã
- ã³ã³ãã¥ãŒã¿ããžã§ã³ã§ã®æŽ»çš:
- ç»åã»åç»ã®èª¬æçæ:
ã¬ãŒã¹ã®ç»åãåç»ãå ¥åãšããŠã銬ã®ç¶æ ãèµ°ãæ¹ãã¬ãŒã¹ã®å±éãªã©ãèªç¶èšèªã§èª¬æããããã¹ããçæã§ããŸããããã«ãããèŠèŠæ å ±ãæ°å€åãã«ããèŠçŽ ãåæã«åãå ¥ããããå¯èœæ§ããããŸãã- ãã«ãã¢ãŒãã«åæ:
LLMã¯ç»åãåç»ã®ç¹åŸŽãšãããã¹ãããŒã¿ïŒã¬ãŒã¹çµæãã³ã¡ã³ããªã©ïŒãçµ±åããŠåæããããšãã§ããŸããããã«ãããããå æ¬çãªéŠ¬ã®ç¶æ ãèœåã®è©äŸ¡ãå¯èœã«ãªããŸãã- ã¢ãããŒã·ã§ã³ã®èªåå:
ã¬ãŒã¹æ åã®ã¢ãããŒã·ã§ã³ïŒäŸïŒéŠ¬ã®åãã®ç¹åŸŽãéšæã®è¡åãªã©ã®ã©ãã«ä»ãïŒãèªååãã倧èŠæš¡ãªããŒã¿ã»ããã®äœæãæ¯æŽã§ããŸãã
- ã²ãŒã çè«ã§ã®æŽ»çš:
- æŠç¥ã®èšèªå:
æ°åŠçã«è¡šçŸãããã²ãŒã çè«ã¢ãã«ã®æŠç¥ããèªç¶èšèªã§èª¬æããããšãã§ããŸããããã«ãããéšæã調æåž«ãçŽæçã«ç解ãããã圢ã§æé©æŠç¥ãæ瀺ã§ããŸãã- ã·ããªãªçæ:
æ§ã ãªã¬ãŒã¹å±éã®ã·ããªãªãèªç¶èšèªã§çæããããããã®ã·ããªãªã«å¯Ÿããæé©æŠç¥ãåæããããšãã§ããŸããããã«ãããããæè»ã§å æ¬çãªæŠç¥ç«æ¡ãå¯èœã«ãªããŸãã- ä»è ã®æå³æšè«:
ã¬ãŒã¹äžã®ä»ã®éšæã®è¡åããããã®æå³ãæŠç¥ãèªç¶èšèªã§æšè«ã説æããããšãã§ããŸããããã«ãããããé«åºŠãªæŠç¥çææ決å®ãå¯èœã«ãªããŸãã
- åé暪æçãªæŽ»çš:
- ç·åçãªã¬ãŒã¹åæ:
çµ±èšããŒã¿ãèŠèŠæ å ±ãæŠç¥åæãçµ±åããå æ¬çãªã¬ãŒã¹äºæž¬ãåæã¬ããŒããèªç¶èšèªã§çæã§ããŸããããã«ãããè€éãªèŠå ãèæ ®ããé«åºŠãªåæçµæãããããããã圢ã§æäŸã§ããŸãã- ã€ã³ã¿ã©ã¯ãã£ããªåææ¯æŽ:
ãŠãŒã¶ãŒããã®è³ªåã«å¯ŸããŠãçµ±èšããŒã¿ãç»åæ å ±ãæŠç¥åæãçµã¿åãããåçãçæã§ããŸããããã«ããããŠãŒã¶ãŒã®é¢å¿ã«å¿ããæè»ãªåæãå¯èœã«ãªããŸãã- ç¶ç¶çåŠç¿ãšç¥èæŽæ°:
æ°ããã¬ãŒã¹çµæãåæèšäºãåžžã«åŠç¿ããããšã§ãææ°ã®åŸåãç¥èŠãåæ ããåæãå¯èœã«ãªããŸãã泚æç¹:
LLMã®åºåã¯åžžã«æ€èšŒãå¿ èŠã§ãããç¹ã«éèŠãªææ決å®ã«ã¯äººéã®å°é家ã«ãã確èªãäžå¯æ¬ ã§ãããŸããå人æ å ±ã®æ±ããå ¬å¹³æ§ã®ç¢ºä¿ãªã©ãå«ççãªé æ ®ãéèŠã«ãªããŸãã
Kelly betting on horse races with uncertainty in probability estimates
1. ã©ããªãã®ïŒ
ãã®è«æã¯ã確çæšå®ã®äžç¢ºå®æ§ãèæ ®ããã±ãªãŒåºæºãããã£ã³ã°ïŒKelly criterion bettingïŒã®æé©åã¢ãã«ãææ¡ããŠããŸããç¹ã«ãå€é ããžã¹ãã£ãã¯ååž°ïŒmultinomial logistic regressionïŒãçšããŠæšå®ããã確çãæã€çžäºæä»çãªçµæã«å¯Ÿããæè³ã®æé©åã«çŠç¹ãåœãŠãŠããŸããèè ã¯ã確çæšå®ã®èª€å·®ããããã£ã³ã°ã®é·æçãªæé·ã«å€§ããªåœ±é¿ãäžããããšã瀺ãããã®äžç¢ºå®æ§ãèæ ®ããã¢ãã«ãæšæºçãªã±ãªãŒåºæºãåæ°ã±ãªãŒæŠç¥ïŒfractional Kelly strategyïŒãããåªããæ§èœãçºæ®ããããšãå®èšŒããŠããŸããææ¡ãããã¢ãã«ã¯ãäž»ã«ç«¶éŠ¬ã®ãããã£ã³ã°ãäŸãšããŠèª¬æãããŠããŸãããããåºç¯ãªæè³èšå®ã«ãé©çšå¯èœã§ããããšã瀺åãããŠããŸãã
2. å è¡ç 究ãæ¯ã¹ãŠã©ãããããïŒ
ãã®ç 究ã®äž»ãªè²¢ç®ã¯ã確çæšå®ã®äžç¢ºå®æ§ãçŽæ¥èæ ®ããã±ãªãŒåºæºãããã£ã³ã°ã®ã¢ãã«ãéçºããããšã§ããåŸæ¥ã®ç 究ã§ã¯ã確çæšå®ã®èª€å·®ã«å¯ŸåŠããããã«åæ°ã±ãªãŒæŠç¥ãçšããããŠããŸããããããã¯ä¿å®çãªã¢ãããŒãã§ãããæœåšçãªå©çãå¶éããå¯èœæ§ããããŸãããæ¬ç 究ã§ã¯ã確çæšå®ã®äžç¢ºå®æ§ãæ瀺çã«ã¢ãã«åããé·æçãªæé·ãæ倧åãã€ã€ãé床ã®ãªã¹ã¯ãåé¿ããã¢ãã«ãææ¡ããŠããŸããç¹ã«ã2ã€ã®çµæãããå Žåã®å³å¯ãªè§£æ³ãšãå€æ°ã®çµæãããå Žåã®è¿äŒŒè§£æ³ãæ瀺ããŠããŸãããŸããã¢ã³ãã«ã«ãç©åãçšããæåŸ å€ã®æšå®ãããã£ã³ã¹å¶çŽïŒchance constraintïŒãçšãããªã¹ã¯ç®¡çãªã©ãæ§ã ãªç¢ºçæé©åææ³ãçµã¿åãããå æ¬çãªã¢ãããŒããææ¡ããŠããŸãã
3. æè¡ãææ³ã®èã¯ã©ãïŒ
æ¬ç 究ã®æ žå¿ã¯ã確çæšå®ã®äžç¢ºå®æ§ãèæ ®ããè€æ°ã®ç¢ºçæé©åã¢ãã«ã®éçºã§ããäž»ãªæè¡çã¢ãããŒãã¯ä»¥äžã®éãã§ãïŒ
- å€é ããžã¹ãã£ãã¯ååž°ãçšãã確çæšå®ãšããã®äžç¢ºå®æ§ã®ã¢ãã«å
- ã¢ã³ãã«ã«ãç©åãçšããæåŸ å¯Ÿæ°å¯ïŒexpected log wealthïŒã®æ倧å
- æ£èŠã¢ãŒã¡ã³ãçæé¢æ°ãçšãã確çã®äžéã®å°åº
- ãã£ã³ã¹å¶çŽãçšãããªã¹ã¯ç®¡ç
- 2ã€ã®çµæãããå Žåã®å³å¯ãªè§£æ³ã®å°åº
- å€æ°ã®çµæãããå Žåã®è¿äŒŒè§£æ³ã®éçº
ãããã®ææ³ãçµã¿åãããããšã§ã確çæšå®ã®äžç¢ºå®æ§ãèæ ®ãã€ã€ãé·æçãªæé·ãæ倧åãããããã£ã³ã°æŠç¥ãå°åºããŠããŸãã4. ã©ããã£ãŠæå¹ã ãšæ€èšŒããïŒ
èè ãã¯ãã·ãã¥ã¬ãŒã·ã§ã³ããŒã¿ãçšããå®èšŒç 究ãè¡ããææ¡ã¢ãã«ã®æå¹æ§ãæ€èšŒããŠããŸããå ·äœçã«ã¯ã以äžã®æé ã§æ€èšŒãè¡ã£ãŠããŸãïŒ
- çã®ç¢ºçååžãçæããããã«åºã¥ããŠååž°ãã©ã¡ãŒã¿ãšãã®äžç¢ºå®æ§ãæš¡æ¬
- ç°ãªãæ¡ä»¶ïŒçµæã®æ°ãææ»çïŒã§4ã€ã®å®éšãèšèš
- åå®éšã§2500åã®è©Šè¡ãå®æœ
- ææ¡ã¢ãã«ãå«ãè€æ°ã®ã¢ãã«ã®æåŸ ç·åçãæ¯èŒ
æ€èšŒã®çµæãææ¡ã¢ãã«ã®äžéšïŒç¹ã«EmcãECCxã¢ãã«ïŒãããã¹ãŠã®å®éšã§æšæºçãªã±ãªãŒåºæºã¢ãã«ãäžåãæ§èœã瀺ããŸããããŸãããã£ã³ã¹å¶çŽãé©åºŠã«äœ¿çšããããšã§ãäžç¢ºå®æ§ã«ããæ倱ãé²ãã€ã€ãé·æçãªæé·ãæ倧åã§ããããšã瀺ãããŸããã5. è°è«ã¯ããïŒ
èè ãã¯ãç 究çµæã«åºã¥ããŠããã€ãã®éèŠãªè°è«ãæèµ·ããŠããŸãïŒ
- 確çæšå®ã®èª€å·®ãææ決å®ã«åãŒã圱é¿ã®é倧æ§ïŒçã®ç¢ºçãšæšå®ç¢ºçãçšããå Žåã®é·æçæé·ã®å€§ããªå·®ã¯ãææ©çåžå Žã§åçãæ倧åããããšãã人ã ã«ãšã£ãŠå€§ããªèª²é¡ã§ããããšã瀺åããŠããŸãã
- ãã£ã³ã¹å¶çŽã®èª¿æŽã®å¿ èŠæ§ïŒãã£ã³ã¹å¶çŽã¯äžç¢ºå®æ§ã«ããæ倱ãé²ãå¹æããããŸãããé床ã«äœ¿çšãããšæœåšçãªå©çãæå¶ããŠããŸãå¯èœæ§ããããŸãããã®ãããåé©çšäºäŸã«å¿ããŠé©åã«èª¿æŽããå¿ èŠããããŸãã
- ææ¡ã¢ãã«ã®äžè¬åã®å¯èœæ§ïŒæ¬ç 究ã§ã¯äž»ã«ç«¶éŠ¬ã®ãããã£ã³ã°ãäŸãšããŠæ±ã£ãŠããŸãããããäžè¬çãªæè³èšå®ãžã®é©çšå¯èœæ§ã瀺åãããŠããŸãã
- å°æ¥ã®ç 究æ¹åïŒèè ãã¯ãå€é ããžã¹ãã£ãã¯ååž°ã®äžç¢ºå®æ§ãçžäºæä»çãªäºè±¡ãè¶ ãããããäžè¬çãªåçèšå®ãžã®ææ³ã®é©çšãä»åŸã®ç 究課é¡ãšããŠææ¡ããŠããŸãã
6. 次ã«èªãã¹ãè«æã¯ïŒ
èè ãã¯çŽæ¥çã«æ¬¡ã«èªãã¹ãè«æãæå®ããŠããŸããããæ¬ç 究ã®èæ¯ãå°æ¥ã®ç 究æ¹åããã以äžã®è«æãåéãé¢é£ããŠãããšèããããŸãïŒ
- Kelly, J.L. (1956). A New Interpretation of Information Rate. - ã±ãªãŒåºæºã®åè«æ
- MacLean, L.C., Ziemba, W.T., & Blazenko, G. (1992). Growth Versus Security in Dynamic Investment Analysis. - åæ°ã±ãªãŒæŠç¥ã«é¢ããç 究
- Baker, R.D., & McHale, I.G. (2013). Optimal Betting Under Parameter Uncertainty: Improving the Kelly Criterion. - ãã©ã¡ãŒã¿äžç¢ºå®æ§äžã§ã®ã±ãªãŒåºæºã®æ¹åã«é¢ããç 究
- 確çæé©åã確çå¶çŽä»ãæé©åã«é¢ããææ°ã®ç 究
- å€é ããžã¹ãã£ãã¯ååž°ã®äžç¢ºå®æ§ã«é¢ããç 究
- éèåžå Žã«ãããæè³æŠç¥ã®æé©åã«é¢ããç 究ãç¹ã«æšå®ã®äžç¢ºå®æ§ãèæ ®ãããã®
ãããã®æç®ãèªãããšã§ãæ¬ç 究ã®èæ¯ãä»åŸã®çºå±æ¹åã«ã€ããŠããæ·±ãç解ãåŸããããšèããããŸãã
ã±ãªãŒåºæºã£ãŠç¥ããªãã£ãã以äžã®ãããªèšäºãèŠã€ããã
ãã以å€ã«ã銬åžã«ãããã±ãªãŒåºæºã£ãŠã®ã¯çµæ§èšäºãããã
ãšããããèããŠã¿ãã
ã±ãªãŒåºæºãšã¯ããããäœãïŒ
ã±ãªãŒåºæºïŒKelly criterionïŒãšã¯
ã±ãªãŒåºæºã¯ãé·æçãªè³ç£æé·ãæ倧åããããã®æè³æŠç¥ãŸãã¯è³ãæ¹ã®æ±ºå®æ¹æ³ã§ãã1956幎ã«ãã«ç 究æã®ãžã§ã³ã»ã©ãªãŒã»ã±ãªãŒã»ãžã¥ãã¢ã«ãã£ãŠèæ¡ãããŸãããäž»ãªç¹åŸŽã¯ä»¥äžã®éãã§ãïŒ
ç®çïŒ
- é·æçãªè³ç£æé·çã®æ倧å
- ç Žç£ãªã¹ã¯ã®æå°å
åºæ¬çãªèãæ¹ïŒ
- åæè³æ©äŒã«å¯ŸããŠãæåŸ å€ãæ£ã§ããå Žåã«ã®ã¿è³ãã
- è³ããéé¡ã¯ãæåŸ å©çãšåçã«åºã¥ããŠæ±ºå®ãã
èšç®æ¹æ³ïŒ
- åçŽãªäºé ååžã®å ŽåïŒf* = (bp - q) / b
- f*: è³ããã¹ãè³ç£ã®å²å
- b: è³ãã«åã£ãå Žåã®çŽå©çç
- p: åã€ç¢ºç
- q: è² ãã確ç (1 - p)
äž»ãªé©çšåéïŒ
- ã®ã£ã³ãã«ïŒç¹ã«ç«¶éŠ¬ããã©ãã¯ãžã£ãã¯ïŒ
- éèæè³ïŒæ ªåŒåžå Žãªã©ïŒ
- æ å ±çè«
é·æïŒ
- çè«çã«æé©ãªé·ææé·çãæäŸ
- ç Žç£ãªã¹ã¯ãæå°å
çæïŒ
- çæçã«ã¯å€§ããªå€åãªã¹ã¯ããã
- æ£ç¢ºãªç¢ºçãšææ»çã®æšå®ãå¿ èŠ
ã±ãªãŒåºæºã¯ãçè«çã«ã¯æé©ãªæŠç¥ãæäŸããŸãããå®è·µã§ã¯ç¢ºçæšå®ã®èª€å·®ãå¿ççèŠå ã«ãããå€ãã®æè³å®¶ã¯ããä¿å®çãªãåæ°ã±ãªãŒæŠç¥ãïŒã±ãªãŒåºæºã®äžéšã®ã¿ãè³ããïŒãæ¡çšããŠããŸããæ¬è«æã¯ããã®ç¢ºçæšå®ã®äžç¢ºå®æ§ã«çŠç¹ãåœãŠãã±ãªãŒåºæºãæ¹è¯ããããšããè©Šã¿ã®äžã€ãšèšããŸãã
ããã競銬ã«çœ®ãæãããšã©ããã䜿ãæ¹ã«ãªãã®ïŒ
競銬ã«ãããã±ãªãŒåºæºã®äœ¿ãæ¹
競銬ã«ã±ãªãŒåºæºãé©çšããå Žåã以äžã®ãããªæé ã§äœ¿çšããŸãïŒ
確çã®æšå®:
- å銬ã®åã€ç¢ºçãæšå®ããŸããããã«ã¯éå»ã®æ瞟ãã³ã³ãã£ã·ã§ã³ãéšæã®è åãªã©ã®èŠçŽ ãèæ ®ããŸãã
- äŸ: ããã¬ãŒã¹ã§ãç¹å®ã®éŠ¬ãåã€ç¢ºçã20%ïŒ0.2ïŒãšæšå®ãããšããŸãã
ãªããºã®ç¢ºèª:
- 銬åžã®ææ»çïŒãªããºïŒã確èªããŸãã
- äŸ: ãã®éŠ¬ã®ãªããºã6.0ïŒ5åïŒã ãšããŸãã
æåŸ å€ã®èšç®:
- æåŸ å€ = (åã€ç¢ºç à ææ»éé¡) - 1
- äŸ: (0.2 à 6.0) - 1 = 0.2 ïŒæ£ã®æåŸ å€ããããŸãïŒ
ã±ãªãŒåºæºã®é©çš:
- f* = (bp - q) / b
- ããã§ãb = ãªã㺠- 1 = 5, p = 0.2, q = 1 - p = 0.8
- f* = (5 Ã 0.2 - 0.8) / 5 = 0.04
è³ãéã®æ±ºå®:
- çŸåšã®è³éã®4%ãè³ããããšãæé©ãšãªããŸãã
- äŸ: ææéã100,000åã®å Žåã4,000åãè³ããŸãã
ãªã¹ã¯ç®¡ç:
- å®è·µã§ã¯ãæšå®èª€å·®ãèæ ®ããŠãèšç®çµæã®ååïŒ2%ïŒã4åã®1ïŒ1%ïŒãè³ãããªã©ãããä¿å®çãªã¢ãããŒããåãããšãäžè¬çã§ãã
ç¶ç¶çãªé©çš:
- åã¬ãŒã¹ãå銬ã«å¯ŸããŠãã®èšç®ãè¡ããæ£ã®æåŸ å€ãããå Žåã«ã®ã¿è³ããŸãã
- ã¬ãŒã¹ããšã«è³éãæŽæ°ããæ°ããè³ãéãèšç®ããŸãã
ã±ãªãŒåºæºã競銬ã«é©çšããéã®æ³šæç¹ïŒ
- 確çæšå®ã®ç²ŸåºŠãéèŠã§ããäžæ£ç¢ºãªæšå®ã¯é·æçã«æ倱ã«ã€ãªããå¯èœæ§ããããŸãã
- åžå Žã®éå¹çæ§ïŒãªããºãšå®éã®ç¢ºçã®ä¹é¢ïŒãèŠåºãã¹ãã«ãå¿ èŠã§ãã
- ææ çãªå€æãæé€ããèšç®çµæã«åºã¥ããŠå·éã«è¡åããããšãéèŠã§ãã
- çæçã«ã¯å€§ããªå€åããããããååãªè³é管çãå¿ èŠã§ãã
æ¬è«æã§ææ¡ãããŠããã¢ãã«ã¯ããããã®èª²é¡ãç¹ã«ç¢ºçæšå®ã®äžç¢ºå®æ§ã«å¯ŸåŠããããšããè©Šã¿ã§ãããããå®å®ããé·æçãªåçãç®æããŠããŸãã
åã€ç¢ºçãã©ãç®åºãã¹ãã ãããïŒ
競銬ã«ãããåã€ç¢ºçã®ç®åºæ¹æ³
åã€ç¢ºçã®æ£ç¢ºãªæšå®ã¯ãã±ãªãŒåºæºãå¹æçã«é©çšããäžã§æ¥µããŠéèŠã§ãã以äžã«ãäžè¬çã«äœ¿çšãããæ¹æ³ãšãã®ç¹åŸŽã瀺ããŸãïŒ
çµ±èšçææ³ïŒ
- å€é ããžã¹ãã£ãã¯ååž°ïŒæ¬è«æã§èšåãããŠããææ³ãè€æ°ã®èŠå ãèæ ®ããŠå銬ã®åçãæšå®ããŸãã
- äŸïŒéå»ã®ã¬ãŒã¹çµæã銬ã®å¹Žéœ¢ãéšæã®æ瞟ãªã©ãå€æ°ãšããŠäœ¿çš
- å©ç¹ïŒè€æ°ã®èŠå ãåæã«èæ ®ã§ãã客芳çãªç¢ºçæšå®ãå¯èœ
- æ¬ ç¹ïŒé©åãªå€æ°éžæãšååãªããŒã¿ãå¿ èŠ
ã¬ãŒãã£ã³ã°æ³ïŒ
- å銬ã«æ°å€ïŒã¬ãŒãã£ã³ã°ïŒãä»ããçžå¯ŸçãªåŒ·ããè©äŸ¡ããŸãã
- äŸïŒã¿ã€ã æ³ïŒéå»ã®ã¿ã€ã ãåºæºã«è©äŸ¡ïŒãéã¿ä»ãã¬ãŒãã£ã³ã°ïŒæ§ã ãªèŠçŽ ã«éã¿ãä»ããŠè©äŸ¡ïŒ
- å©ç¹ïŒçŽæçã§ç解ãããããæ¯èŒçç°¡åã«å®è£ å¯èœ
- æ¬ ç¹ïŒã¬ãŒãã£ã³ã°ã®æŽæ°æ¹æ³ãéã¿ä»ãã«äž»èŠ³ãå ¥ãå¯èœæ§ããã
ã·ãã¥ã¬ãŒã·ã§ã³æ³ïŒ
- ã¢ã³ãã«ã«ãã·ãã¥ã¬ãŒã·ã§ã³ãªã©ãçšããŠãã¬ãŒã¹ãäœåºŠãä»®æ³çã«å®è¡ããŸãã
- äŸïŒå銬ã®èœåå€ã確çååžãšããŠè¡šçŸããå€æ°åã®ã·ãã¥ã¬ãŒã·ã§ã³ãå®è¡
- å©ç¹ïŒè€éãªçžäºäœçšãäžç¢ºå®æ§ãèæ ®å¯èœ
- æ¬ ç¹ïŒé©åãªã¢ãã«èšèšãšèšç®ãªãœãŒã¹ãå¿ èŠ
æ©æ¢°åŠç¿ã¢ãããŒãïŒ
- ãã¥ãŒã©ã«ãããã¯ãŒã¯ã決å®æšãªã©ã®ææ³ãçšããŠãéå»ã®ããŒã¿ããåŠç¿ããŸãã
- äŸïŒæ·±å±€åŠç¿ãçšããŠãç»åããŒã¿ïŒéŠ¬ã®ç¶æ ïŒãå«ããç·åçãªè©äŸ¡
- å©ç¹ïŒè€éãªãã¿ãŒã³ãéç·åœ¢ãªé¢ä¿æ§ãæããããå¯èœæ§ããã
- æ¬ ç¹ïŒå€§éã®è³ªã®é«ãããŒã¿ãå¿ èŠã§ã解éãé£ããå Žåããã
ãšãã¹ããŒãã®å€æïŒ
- 競銬å°é家ã®åæãäºæ³ãæ°å€åããŸãã
- äŸïŒè€æ°ã®å°é家ã®äºæ³ãéçŽãã確çã«å€æ
- å©ç¹ïŒå®éåãé£ããèŠå ïŒéŠ¬ã®æ°æ§ãªã©ïŒãèæ ®å¯èœ
- æ¬ ç¹ïŒäž»èŠ³çã§ããããã€ã¢ã¹ãå ¥ãå¯èœæ§ããã
ãã€ããªããã¢ãããŒãïŒ
- äžèšã®æ¹æ³ãçµã¿åãããŠäœ¿çšããŸãã
- äŸïŒçµ±èšã¢ãã«ã®çµæãšãšãã¹ããŒãã®å€æãçµ±å
- å©ç¹ïŒåææ³ã®é·æã掻ããããã粟床ã®é«ãæšå®ãå¯èœ
- æ¬ ç¹ïŒè€éã«ãªãããããåææ³ã®éã¿ä»ããé£ãã
å®è·µçãªã¢ãããŒãïŒ
- è€æ°ã®ææ³ã䜵çšããçµæãæ¯èŒæ€èšŒããããšãæšå¥šãããŸãã
- ç¶ç¶çãªã¢ãã«ã®è©äŸ¡ãšæ¹åãéèŠã§ãã
- åžå Žã®ãªããºãšã®æ¯èŒãè¡ããèªèº«ã®æšå®ã®åŠ¥åœæ§ã確èªããŸãã
- äžç¢ºå®æ§ãèæ ®ããæšå®ç¢ºçã«å¯ŸããŠé©åãªä¿¡é Œåºéãèšå®ããããšãæå¹ã§ãã
æ¬è«æã§ææ¡ãããŠããã¢ãã«ã¯ããããã®ç¢ºçæšå®æ¹æ³ã«å åšããäžç¢ºå®æ§ãæ瀺çã«èæ ®ããããšã§ãããå ç¢ãªææ決å®ãå¯èœã«ããããšããè©Šã¿ãšèšããŸãã
ãªãã»ã©ãææ°ãšã¯çžæ§ãè¯ãããã§ã¯ãããããããããåã€ç¢ºçãç®åºããããšèªäœãé£ããæ°ã¯ããã
ã§ããã®è«æã ãšãã®åçãç®åºããã®ã¯äžå¯èœãªã®ã§ãäžå®ã®äžç¢ºå®æ§ãèæ ®ã«å ¥ããŠãåççã»é·æçã«ãã£ãŠãããšããã¢ãã«ã®ããã«æããããã®ããã®ææ決å®ããã»ã¹ã¯ãããªæãã
è«æã§ææ¡ãããŠããææ決å®ããã»ã¹
åºæ¬çãªæµãïŒ
- 確çæšå® â äžç¢ºå®æ§ã®èæ ® â æé©å â è³ãé決å®
確çæšå®ïŒå€é ããžã¹ãã£ãã¯ååž°ïŒïŒ
- å ¥åïŒéŠ¬ã®ç¹åŸŽïŒéå»ã®æ瞟ãéšæãéŠ¬å Žç¶æ ãªã©ïŒ
- åºåïŒå銬ã®æšå®åç Ï^h
- ç¹åŸŽïŒååž°ä¿æ° β ã®æšå®å€ãšå ±åæ£è¡å Σ ãç®åº
äžç¢ºå®æ§ã®èæ ®ïŒ
- æšå®ããã β ãå¹³åãΣ ãå ±åæ£è¡åãšããæ£èŠååžãä»®å®
- ããã«ãããåçæšå®ã®äžç¢ºå®æ§ã確çååžãšããŠè¡šçŸ
æé©åã¢ãã«ã®éžæïŒ
- è«æã§ã¯è€æ°ã®ã¢ãã«ãææ¡ïŒElb, Emc, CC, ECC ãªã©ïŒ
- äŸïŒEmcïŒã¢ã³ãã«ã«ãæåŸ å€æ倧åïŒ
- å€æ°ã® β ãµã³ãã«ãçæããå¹³åæåŸ å¯Ÿæ°å¯ãæ倧å
- äŸïŒCCïŒãã£ã³ã¹å¶çŽä»ãæé©åïŒ
- äžå®ç¢ºçïŒ1-αïŒã§æå°éã®åçãä¿èšŒ
æé©ååé¡ã®è§£æ±ºïŒ
- ç®çïŒæåŸ å¯Ÿæ°å¯ã®æ倧å
- å¶çŽïŒç·è³ãé †çŸåšã®å¯ãåè³ãé ⥠0
- çµæïŒå銬ãžã®æé©è³ãéæ¯ç x^h ãç®åº
è³ãéã®æ±ºå®ïŒ
- çŸåšã®è³éã«æé©è³ãéæ¯çãä¹ããŠå®éã®è³ãéã決å®
- äŸïŒè³é100äžåãx^1 = 0.05 ãªãã1å·éŠ¬ã«5äžåè³ãã
ãªã¹ã¯èª¿æŽïŒãªãã·ã§ã³ïŒïŒ
- ãã£ã³ã¹å¶çŽã®Î±å€ã調æŽããŠãªã¹ã¯èš±å®¹åºŠãå€æŽå¯èœ
- αå€ãå°ããããã»ã©ãªã¹ã¯åé¿çã«ãªã
å®è·µçãªèæ ®äºé ïŒ
- å šãŠã®ã¬ãŒã¹ã§è³ããã®ã§ã¯ãªããæå©ãªç¶æ³ã®ã¿éžæ
- åžå Žã®ãªããºãšæ¯èŒããæåŸ å€ãæ£ã®å Žåã®ã¿è³ãã
- è³é管çãèæ ®ããäžåã®è³ãéãç·è³éã®äžå®å²åãè¶ ããªããã調æŽ
ç¶ç¶çãªè©äŸ¡ãšèª¿æŽïŒ
- å®éã®çµæãšäºæž¬ãæ¯èŒããã¢ãã«ã®æ§èœãè©äŸ¡
- å¿ èŠã«å¿ããŠã¢ãã«ã®ãã©ã¡ãŒã¿ãæ§é ã調æŽ
ãã®ææ決å®ããã»ã¹ã®ç¹åŸŽã¯ã確çæšå®ã®äžç¢ºå®æ§ãæ瀺çã«èæ ®ããŠããç¹ã§ããããã«ãããé床ã«æ¥œèŠ³çãªè³ããæå¶ããé·æçã«ããå®å®ããæé·ãç®æããŠããŸãããŸãããªã¹ã¯èš±å®¹åºŠã®èª¿æŽãå¯èœãªæè»ãªãã¬ãŒã ã¯ãŒã¯ãšãªã£ãŠããŸãã
äºæ³ãšããããã¯éŠ¬åžã®è²·ãæ¹ã®èŠç¹ãšããŠæããã»ããè¯ããããªæ°ã¯ããã
A Horse with no Labels: Self-Supervised Horse Pose Estimation from Unlabelled Images and Synthetic Prior
ã©ããªãã®ïŒ
ãã®è«æã¯ãã©ãã«ä»ããããŠããªãç»åãšå°éã®åæ2DããŒãºããŒã¿ã®ã¿ã䜿çšããŠã銬ã®3DããŒãºãæšå®ããèªå·±æåž«ããææ³ãææ¡ããŠããŸããåŸæ¥ã®åç©ããŒãºæšå®ææ³ã§ã¯ãèšç·ŽããŒã¿ã«ã©ãã«ä»ããå¿ èŠã§ããããæ¬ææ³ã§ã¯ãããäžèŠãšããŠããŸããäž»ãªç¹åŸŽã¯ä»¥äžã®éãã§ãïŒ
- ã©ãã«ãªãç»åãšåæ2DããŒãºã®ã¿ã䜿çš
- 3Dãš2DããŒãºãåæã«åŠç¿
- è€éãª3Dåç©ã¢ãã«ãå¿ èŠãšããªã
- YouTubeããåéãã銬ã®ç»åã§èšç·Ž
- åæããŒã¿ããçæãã2DããŒãºãäºåç¥èãšããŠäœ¿çš
ãã®ææ³ã«ãããã©ãã«ä»ãããŒã¿ã®äžè¶³ãšãã課é¡ãå æããåç©ããŒãºæšå®ã®æ°ããªå¯èœæ§ã瀺ããŠããŸãã
å è¡ç 究ãšæ¯ã¹ãŠã©ãããããïŒ
æ¬ç 究ã®é©æ°æ§ã¯ã以äžã®ç¹ã«ãããŸãïŒ
- æå°éã®åææ¡ä»¶ïŒå€ãã®å è¡ç 究ãåæç»åãäºåèšç·Žã¢ãã«ãå¿ èŠãšããŸãããæ¬ææ³ã¯ã©ãã«ãªãç»åãšå°éã®åæ2DããŒãºã®ã¿ã§åäœããŸãã
- ãšã³ãããŒãšã³ãã®åŠç¿ïŒ2Dãš3DããŒãºãåæã«åŠç¿ããäžéè¡šçŸãç²åŸããŸãã
- å¹ åºãé©çšå¯èœæ§ïŒæå°éã®èŠä»¶ã«ãããæ§ã ãªåç©çš®ã«å®¹æã«é©çšã§ããå¯èœæ§ããããŸãã
- èªå·±æåž«ããåŠç¿ïŒã©ãã«ä»ãããŒã¿ãå¿ èŠãšããã幟äœåŠçäžè²«æ§ãå©çšããŠ3DããŒãºãåŠç¿ããŸãã
- ãã¡ã€ã³é©å¿ã®ç°¡çŽ åïŒåæããŒã¿ãšå®ããŒã¿ã®ã®ã£ãããåããããã®è€éãªåŠçãå¿ èŠãšããŸããã
ãããã®ç¹åŸŽã«ãããæ¬ææ³ã¯åç©ããŒãºæšå®ã®åéã«æ°ããªã¢ãããŒãããããããŠããŸãã
æè¡ãææ³ã®èã¯ã©ãïŒ
æ¬ææ³ã®æ žå¿ã¯ä»¥äžã®èŠçŽ ã«ãããŸãïŒ
- ç»åãã3DããŒãºãžã®ãããã³ã°ïŒCNNïŒç³ã¿èŸŒã¿ãã¥ãŒã©ã«ãããã¯ãŒã¯ïŒãçšããŠãå ¥åç»åããäžéçãªã¹ã±ã«ãã³ç»åã2DããŒãºããããŠæçµçã«3DããŒãºãžãšå€æããŸãã
- èªå·±æåž«ããåŠç¿ïŒ3DããŒãºã®å転ãšæ圱ã®ã«ãŒããå©çšããŠã幟äœåŠçäžè²«æ§ãä¿èšŒããŸãã
- æµå¯ŸçåŠç¿ïŒå€å¥åšãããã¯ãŒã¯ãšåæ2DããŒãºã®äºåååžãçšããŠãçæãããã¹ã±ã«ãã³ç»åã®çŸå®æ§ãé«ããŸãã
- æ倱é¢æ°ïŒ2DããŒãºã3DããŒãºãå転åŸã®3DããŒãºã®äžè²«æ§ãä¿ã€ããã®æ倱é¢æ°ãçµã¿åãããŠäœ¿çšããŸãã
- åæããŒã¿ã®æŽ»çšïŒCADã¢ãã«ããçæãã2DããŒãºãäºåç¥èãšããŠå©çšããå®ããŒã¿ãšã®ã®ã£ãããæå°éã«æããŸãã
ãããã®èŠçŽ ãçµã¿åãããããšã§ãã©ãã«ãªãããŒã¿ããã®å¹æçãª3DããŒãºæšå®ãå®çŸããŠããŸãã
ã©ããã£ãŠæå¹ã ãšæ€èšŒããïŒ
æ¬ç 究ã§ã¯ã以äžã®æ¹æ³ã§ææ¡ææ³ã®æå¹æ§ãæ€èšŒããŠããŸãïŒ
- 2DããŒãºæšå®ã®å®éçè©äŸ¡ïŒWeizmannããŒã¿ã»ããã®æåã¢ãããŒã·ã§ã³ãçšããŠãPCK@0.05ã¡ããªãã¯ã§2DããŒãºæšå®ã®ç²ŸåºŠãè©äŸ¡ããŸããã
- 3DããŒãºæšå®ã®å®æ§çè©äŸ¡ïŒèšç·ŽããŒã¿ãšã¯ç°ãªãWeizmannããŒã¿ã»ããã®ç»åã䜿çšããæšå®ããã3DããŒãºãèŠèŠçã«è©äŸ¡ããŸããã
- ä»ææ³ãšã®æ¯èŒïŒåæ§ã®æ¡ä»¶äžã§åäœãã2DããŒãºæšå®ææ³ãšæ¯èŒããæ¬ææ³ã®åªäœæ§ã瀺ããŸããã
- æ±åæ§èœã®æ€èšŒïŒéŠ¬ã§èšç·Žããã¢ãã«ãã·ããŠãã®ç»åã«é©çšããé¡äŒŒããåç©çš®ãžã®æ±åèœåã確èªããŸããã
- åæããŒã¿ã§ã®èšç·Žå®éšïŒå®å šã«åæçãªããŒã¿ïŒã·ããŠãã®åæç»åãšéŠ¬ã®åæ2DããŒãºïŒã§èšç·Žããã¢ãã«ãšãå®ç»åã§èšç·Žããã¢ãã«ã®æ§èœãæ¯èŒããŸããã
ãããã®è©äŸ¡ã«ãããææ¡ææ³ã®æå¹æ§ãšæ±çšæ§ã瀺ãããŠããŸãã
è°è«ã¯ããïŒ
è«æã§ã¯ä»¥äžã®ç¹ã«ã€ããŠè°è«ããªãããŠããŸãïŒ
- æ¹åã®äœå°ïŒèè ãã¯çŸåšã®çµæã«ã¯ãããªãæ¹åã®äœå°ããããšèªèããŠããŸãã
- æéçæ å ±ã®æŽ»çšïŒå°æ¥ã®ç 究æ¹åãšããŠãã¢ãããŒãã«æéçæ å ±ãçµã¿èŸŒãããšãææ¡ããŠããŸããããã«ãããé£ç¶ãããã¬ãŒã éã®äžè²«æ§ãåäžããå¯èœæ§ããããŸãã
- ãã¡ã€ã³é©å¿ïŒå°éã®å®ããŒã¿ãçšãããã¡ã€ã³ãã¥ãŒãã³ã°ã«ãããåæããŒã¿ãšå®ããŒã¿ã®ãã¡ã€ã³ã®ã£ãããããã«çž®å°ã§ããå¯èœæ§ã瀺åãããŠããŸãã
- 2DããŒãºæšå®ã®åœ±é¿ïŒ2DããŒãºæšå®ã®ç²ŸåºŠã3DããŒãºæšå®ã®ç²ŸåºŠã«åœ±é¿ãäžããããšãææãããŠããŸãããã ããäžæ£ç¢ºãª2DããŒãºæšå®ããã劥åœãª3DããŒãºãå埩ã§ããå Žåãããããšã瀺ãããŠããŸãã
- ããŒã¿èŠä»¶ã®æå°åïŒææ¡ææ³ã¯èšç·ŽããŒã¿ã®èŠä»¶ãæå°éã«æããŠãããããã«ããå€ãã®ã©ãã«ãªãããŒã¿ã»ããã«é©çšå¯èœã§ããããšãè°è«ãããŠããŸãã
ãããã®è°è«ç¹ã¯ãæ¬ç 究ã®éçãšå°æ¥ã®ç 究æ¹åã瀺ããŠããŸãã
次ã«èªãã¹ãè«æã¯ïŒ
æ¬è«æã®å 容ãæ·±ãç解ããé¢é£ããç 究ãæ¢æ±ããããã«ã以äžã®è«æãèªãããšããå§ãããŸãïŒ
- Sosa, J., & Hogg, D. (2023). "Self-supervised 3D Human Pose Estimation from a Single Image." CVPR 2023.
- æ¬ç 究ã®åºç€ãšãªã£ã人éã®ããŒãºæšå®ææ³ã«ã€ããŠè©³ããç¥ãããšãã§ããŸãã
- Chen, C. H., et al. (2019). "Unsupervised 3D Pose Estimation with Geometric Self-Supervision." CVPR 2019.
- 幟äœåŠçèªå·±æåž«ããåŠç¿ã®æŠå¿µãããæ·±ãç解ã§ããŸãã
- Mu, J., et al. (2020). "Learning from Synthetic Animals." CVPR 2020.
- åç©ããŒãºæšå®ã«ãããåæããŒã¿ã®æŽ»çšã«ã€ããŠè©³ããåŠã¹ãŸãã
- Li, C., & Lee, G. H. (2021). "From Synthetic to Real: Unsupervised Domain Adaptation for Animal Pose Estimation." CVPR 2021.
- åç©ããŒãºæšå®ã«ããããã¡ã€ã³é©å¿ã®ææ³ãç解ããã®ã«åœ¹ç«ã¡ãŸãã
- Zuffi, S., et al. (2019). "Three-D Safari: Learning to Estimate Zebra Pose, Shape, and Texture from Images 'in the Wild'." ICCV 2019.
- éçåç©ã®3DããŒãºæšå®ã«é¢ããå é§çãªç 究ãåŠã¶ããšãã§ããŸãã
ãããã®è«æãèªãããšã§ãåç©ã®3DããŒãºæšå®ãšé¢é£ããæè¡ã«ã€ããŠããå æ¬çãªç解ãåŸãããã§ãããã
銬äœåçãšããããã¯ç»åãšãããéªšæ Œæšå®ãããã«ããããæãã«åãæ§å³ã§çµ±äžããã®ã¯é£ãããã©ãæšå®ãšã¯ãããããªãã°äœãããã®ç»åããŒã¿ãããã°ããããããªãã®ãããããªããªãã
ããããã°mmpose觊ã£ãŠãªãã»ã»ã»ç°å¢æ§ç¯ã ãã§ããªãé¢åã ã£ãã®ã§ããã§å¿æãããã ããªã»ã»ã»
hSMAL: Detailed Horse Shape and Pose Reconstruction for Motion Pattern Recognition
1. ã©ããªãã®ïŒ
ãã®è«æã¯ã銬ã®3D圢ç¶ãšå§¿å¢ã®è©³çŽ°ãªåæ§æãããã³ãã®ããŒã¿ãçšããè·è¡ïŒã¯ããïŒæ€åºã«é¢ããç 究ã玹ä»ããŠããŸããèè ãã¯ãSMALïŒShape, Motion and Appearance for AnimalsïŒã¢ãã«ã銬çšã«ç¹åããããhSMALãã¢ãã«ãéçºããŸããããã®ã¢ãã«ã¯ã37åã®éŠ¬ã®ããã¡ãããåŠç¿ããæ°ãããã³ãã¬ãŒããéªšæ Œã圢ç¶ç©ºéãå®çŸ©ããŠããŸããhSMALã¢ãã«ã¯ã3Dã¢ãŒã·ã§ã³ãã£ããã£ãŒããŒã¿ãšç»åãã銬ãåæ§æãã粟床ããã¹ããããããã«åç»ããã®è·è¡æ€åºåé¡ã«ãé©çšãããŸãããè·è¡æ€åºã«ã¯ã3DããŒãºããŒã¿ã«å¯ŸããŠèšç·ŽãããST-GCNïŒSpatial Temporal Graph Convolutional NetworkïŒãããã¯ãŒã¯ã䜿çšãããŠããŸãããã®ç 究ã¯ã銬ã®åäœãã¿ãŒã³èªèã¿ã¹ã¯ã«ããã3Dã¢ãã«ããŒã¹ã®è¡šçŸã®å©ç¹ã瀺ããŠããŸãã
2. å è¡ç 究ãšæ¯ã¹ãŠã©ãããããïŒ
ãã®ç 究ã®äž»ãªé©æ°ç¹ã¯ã銬ã«ç¹åãã3Dã¢ãã«ïŒhSMALïŒã®éçºãšããããçšããè·è¡æ€åºãžã®å¿çšã§ããåŸæ¥ã®åç©ã®3Dåæ§æç 究ã¯ãäž»ã«çš®éã®è¡šçŸã«çŠç¹ãåœãŠãŠããŸããããæ¬ç 究ã¯çš®å ã®è©³çŽ°ãªåæãå¯èœã«ããŠããŸããè·è¡æ€åºã«é¢ããŠã¯ãåŸæ¥ã®æ¹æ³ãå é床èšããžã£ã€ãã¹ã³ãŒããªã©ã®ãŠã§ã¢ã©ãã«ã»ã³ãµãŒã䜿çšããŠããã®ã«å¯Ÿãããã®ç 究ã§ã¯éæ¥è§Šçãªãããªè§£æã¢ãããŒããææ¡ããŠããŸããããã«ã人éã®è¡åèªèã§æåãåããŠããST-GCNã銬ã®åäœãã¿ãŒã³èªèã«é©çšããç¹ãæ°ããã§ããããã«ããã銬ã®è©³çŽ°ãª3D圢ç¶ãšå§¿å¢ã®æ å ±ã掻çšãããããé«åºŠãªåäœãã¿ãŒã³åæãå¯èœã«ãªããŸããã
3. æè¡ãææ³ã®èã¯ã©ãïŒ
æ¬ç 究ã®æ žå¿ã¯ã以äžã®3ã€ã®æè¡çã€ãããŒã·ã§ã³ã«ãããŸãïŒ
- hSMALã¢ãã«ïŒSMALã¢ãã«ãåºã«ã銬å°çšã®3Dçµ±èšç圢ç¶ã¢ãã«ãéçºãæ°ãã銬ã®ãã³ãã¬ãŒããéªšæ Œã37åã®éŠ¬ã®ããã¡ãããåŠç¿ãã圢ç¶ç©ºéãå®çŸ©ããŠããŸãã
- ã¢ãã«ãã£ããã£ã³ã°ïŒç»åã2DããŒãã€ã³ãã3Dã¢ãŒã·ã§ã³ãã£ããã£ãŒããŒã¿ã«å¯ŸããŠhSMALã¢ãã«ããã£ããã£ã³ã°ããææ³ãéçºããšãã«ã®ãŒç®çé¢æ°ãæå°åããããšã§ãäœã®åœ¢ç¶ãããŒãºãå šäœçãªå¹³è¡ç§»åãæšå®ããŸãã
- è·è¡æ€åºïŒST-GCNãçšããŠãhSMALã¢ãã«ããæœåºããããŒãºãã©ã¡ãŒã¿ãé¢ç¯åº§æšã®ã·ãŒã±ã³ã¹ããè·è¡ãåé¡ã銬ã®æ§é ã«åºã¥ããã°ã©ãã空éé åã§åœ¢æããé£ç¶ãã¬ãŒã éã§åãé¢ç¯ãæ¥ç¶ããããšã§æéçãªæ å ±ãæããŠããŸãã
ãããã®æè¡ã«ããã2Dç»åãã銬ã®è©³çŽ°ãª3D圢ç¶ãšå§¿å¢ãåæ§æãããã®æ å ±ãçšããŠè€éãªåäœãã¿ãŒã³ãåæããããšãå¯èœã«ãªããŸããã4. ã©ããã£ãŠæå¹ã ãšæ€èšŒããïŒ
ç 究ã®æå¹æ§ã¯äž»ã«ä»¥äžã®2ã€ã®å®éšã§æ€èšŒãããŠããŸãïŒ
- hSMALã¢ãã«ã®ç²ŸåºŠè©äŸ¡ïŒ
- 20ã®ä»£è¡šçãªãã¬ãŒã ãéžã³ãhSMALãšSMALã¢ãã«ã®åæ§æ粟床ãæ¯èŒã
- ã¢ãŒã·ã§ã³ãã£ããã£ãŒããŒã¿ãšã¢ãã«ã®åæ§æãããç¹ãšã®å¹³åè·é¢ãèšç®ã
- çµæïŒhSMALã®å¹³å誀差ã¯çŽ7cmãSMALã¯çŽ14cmã§ãhSMALã®åªäœæ§ã瀺ããŸããã
- ããã«ãç°ãªãå ¥åïŒ3Dã¢ãŒã·ã§ã³ãã£ããã£ãŒããŒã¿ã2DããŒãã€ã³ããã·ã«ãšããïŒã§ã®åæ§æ粟床ãIOUïŒIntersection over UnionïŒã§è©äŸ¡ã
- è·è¡æ€åºã®æ§èœè©äŸ¡ïŒ
- 7é ã®éŠ¬ã®è¢«éšè ããŒã¿ãçšããŠã5ã¯ã©ã¹ïŒæ£åžžãåè¢å·Šå³ãåŸè¢å·Šå³ã®è·è¡ïŒã®åé¡ã¿ã¹ã¯ãå®æœã
- ST-GCNãçšããŠãçã®3Dã¢ãŒã·ã§ã³ãã£ããã£ãŒããŒã¿ãhSMALã®ããŒãºãã©ã¡ãŒã¿ãé¢ç¯äœçœ®ããããã§åé¡åšãèšç·Žã
- 亀差æ€èšŒæ³ãçšããŠè©äŸ¡ããããŒã¿æ¡åŒµã®å¹æãæ€èšŒã
- çµæïŒhSMALã¢ãã«ãçšããå Žåãã¢ãŒã·ã§ã³ãã£ããã£ãŒããŒã¿ãšåçããã以äžã®åé¡æ§èœã瀺ããŸããã
ãããã®å®éšçµæã¯ãhSMALã¢ãã«ã®ç²ŸåºŠãšããããçšããè·è¡æ€åºææ³ã®æå¹æ§ãå®èšŒããŠããŸãã5. è°è«ã¯ããïŒ
èè ãã¯ä»¥äžã®ç¹ã«ã€ããŠè°è«ããŠããŸãïŒ
- hSMALã¢ãã«ã®ç²ŸåºŠïŒå®éšçµæãããhSMALã¢ãã«ãSMALã¢ãã«ãããå®éã®éŠ¬ã®åœ¢ç¶ãããæ£ç¢ºã«åæ§æã§ããããšã瀺ãããŸãããããã¯ãçš®ç¹ç°çãªã¢ããªã³ã°ã®éèŠæ§ã瀺åããŠããŸãã
- è·è¡æ€åºã«ãããè¡šçŸã®éèŠæ§ïŒhSMALã¢ãã«ããæœåºããããŒãºããŒã¿ããçã®ã¢ãŒã·ã§ã³ãã£ããã£ãŒããŒã¿ãšåçããã以äžã®åé¡æ§èœã瀺ããããšããã3Dã¢ãã«ããŒã¹ã®è¡šçŸãåäœãã¿ãŒã³èªèã¿ã¹ã¯ã«é©ããŠããããšã瀺åãããŸããã
- ã¢ãã«ã®æœåšçãªå©ç¹ïŒèè ãã¯ãhSMALã¢ãã«ãåã«ã¢ãŒã·ã§ã³ãã£ããã£ãŒããŒã¿ã®æ å ±ãä¿åããã ãã§ãªãã銬ã®äœã®è§£ååŠçãªäºåç¥èãå¶çŽã«é¢ããæ å ±ãä»å ããŠããå¯èœæ§ãææããŠããŸãã
- å°æ¥ã®ç 究æ¹åïŒèè ãã¯ãã¡ãã·ã¥äžã®ããŒã«ãŒéžæã®æé©åãå®éã®éŠ¬ã®3Dã¹ãã£ã³ããŒã¿ãçšãã圢ç¶ç©ºéã®æ¹åããããŠã¢ãŒã·ã§ã³ãã£ããã£ãŒããŒã¿ãå¿ èŠãšããªããããå¶çŽã®å°ãªãèšå®ã§ã®å¿çšãä»åŸã®èª²é¡ãšããŠæããŠããŸãã
ãããã®è°è«ã¯ãæ¬ç 究ã®ææãšéçãæ確ã«ããä»åŸã®ç 究ã®æ¹åæ§ã瀺ããŠããŸãã6. 次ã«èªãã¹ãè«æã¯ïŒ
æ¬ç 究ã®ç解ãæ·±ããé¢é£åéã®ææ°ååãææ¡ããããã«ã以äžã®è«æãèªãããšããå§ãããŸãïŒ
- Zuffi et al. (2017) "3D Menagerie: Modeling the 3D shape and pose of animals" - SMALã¢ãã«ã®åè«æã§ãæ¬ç 究ã®åºç€ãšãªãåç©ã®3Dã¢ããªã³ã°ææ³ãç解ããã®ã«åœ¹ç«ã¡ãŸãã
- Biggs et al. (2018) "Creatures great and SMAL: Recovering the shape and motion of animals from video" - åç»ããã®åç©ã®åœ¢ç¶ãšåãã®åŸ©å ã«é¢ããç 究ã§ãæ¬ç 究ãšæ¯èŒããããšã§ã銬ç¹æã®èª²é¡ã解決çãç解ã§ããŸãã
- Yan et al. (2018) "Spatial temporal graph convolutional networks for skeleton-based action recognition" - æ¬ç 究ã§äœ¿çšãããŠããST-GCNã®åè«æã§ãåäœèªèã®ããã®æè¡çèæ¯ãç解ããã®ã«åœ¹ç«ã¡ãŸãã
- Kearney et al. (2020) "RGBD-Dog: Predicting canine pose from RGBD sensors" - ç¬ã®3DããŒãºæšå®ã«é¢ããææ°ã®ç 究ã§ãä»ã®åç©çš®ã§ã®é¡äŒŒã¢ãããŒããç解ããã®ã«æçšã§ãã
- Rhodin et al. (2018) "Vertical movement symmetry of the withers in horses with induced forelimb and hindlimb lameness at trot" - æ¬ç 究ã§äœ¿çšãããããŒã¿ã»ããã«é¢é£ããè«æã§ã銬ã®è·è¡ã®çäœååŠçåŽé¢ãç解ããã®ã«åœ¹ç«ã¡ãŸãã
ãããã®è«æãèªãããšã§ãåç©ã®3Dã¢ããªã³ã°ã姿å¢æšå®ããããŠåäœãã¿ãŒã³èªèã®åéã«ãããææ°ã®ç 究ååãææ¡ããããšãã§ããŸãã
ãããè·è¡æ€åºãšãç±ããªãããããå®éã«åœ¹ã«ç«ã¡ãããªãã®ãæ¬æ¥æè¡ã®ããã¹ã姿ã ããã
ããããã°ããã®åãã£ãŠãã以äžã®çªçµ
çªçµäžã§ã調æåã«éŠ¬ã®è¶³ã®éã³ãè¯ããªãã®ã§èª¿æäŒãŸãããã¿ããã®ã玹ä»ãããŠããã®ã ãã©ãèªåã¯äœãã©ãæªãã®ããã£ã±ãããããªãã£ããããã¯ãã£ã±ãéããªãšæãã®ã ãã©ãããããæè¡ã掻çšããŠãã軜ããã¡ã«çºèŠã§ãããšè¯ããªãšæãããã
Emotion Recognition in Horses with Convolutional Neural Networks
1. ã©ããªãã®ïŒ
ãã®è«æã¯ãæ©æ¢°åŠç¿ãçšããŠéŠ¬ã®ææ ãèªèããã·ã¹ãã ã®éçºã«ã€ããŠèª¬æããŠããŸããã·ã¹ãã ã¯2ã€ã®äž»èŠãªéšåããæ§æãããŠããŸãïŒ
- 銬ã®é ãšéŠãç»åããæ€åºããæ€åºåš
- æ€åºãããé åãã銬ã®ææ ãäºæž¬ããã¢ãã«
ææ ã¯ãèŠæããäžæ©å«ãã奜å¥å¿ãããªã©ãã¯ã¹ãã®4ã€ã«åé¡ãããŸãããããã®ææ ã¯ã銬ã®è¡ååŠçç 究ã«åºã¥ããŠå®çŸ©ãããŠãããç®ãè³ã錻ãéŠã®äœçœ®ãªã©ã®èŠèŠçãªæãããã䜿çšããŠããŸããã·ã¹ãã ã¯çç£æ¥ãç£å»åŠåéã§ã®åç©çŠç¥ã®åäžã«è²¢ç®ããå¯èœæ§ããããŸãã2. å è¡ç 究ãšæ¯ã¹ãŠã©ãããããïŒ
ãã®ç 究ã®æ°èŠæ§ã¯ãæ©æ¢°åŠç¿ãçšããŠåç©ãç¹ã«éŠ¬ã®ææ ãèªåçã«èªèããããšããç¹ã«ãããŸããå è¡ç 究ã§ã¯ïŒ
- 人éã®ææ èªèã·ã¹ãã ã®éçºïŒäŸïŒAffectiva瀟ïŒ
- åç©ã®ææ ã«é¢ããç¥çµç§åŠçã»å¿çåŠçç 究
- åç©ã®è¡šæ ã人éã解æããããŒã«ïŒäŸïŒAnimalFacsïŒ
ãªã©ããããŸããããèªåŸçãªã·ã¹ãã ã«ããåç©ã®ææ èªèã¯ããŸãæ¢æ±ãããŠããŸããã§ãããç¬ã®ææ ãäºæž¬ããç 究ã¯ãããŸãããã銬ã«ç¹åãããã®ã¯ãããŸããã§ããããã®ç 究ã¯ãçç£åç©ã®ææ ç 究ãšããæ°ããåéãéæããŠããŸãã3. æè¡ãææ³ã®èã¯ã©ãïŒ
ãã®ã·ã¹ãã ã®æ žãšãªãæè¡ã¯ä»¥äžã®2ã€ã§ãïŒ
- 銬ã®æ€åºåšïŒFaster R-CNNïŒRegion-based Convolutional Neural NetworkïŒã¢ãŒããã¯ãã£ã䜿çšãç»åãã銬ã®é ãšéŠã®é åãé«ç²ŸåºŠã§æ€åºããŸãã
- ææ äºæž¬ã¢ãã«ïŒVGG16ãããŒã¹ãšããç³ã¿èŸŒã¿ãã¥ãŒã©ã«ãããã¯ãŒã¯ïŒCNNïŒã䜿çšãæ€åºãããé åãã4ã€ã®ææ ã«ããŽãªãäºæž¬ããŸãã
éèŠãªãã€ã³ãã¯ã銬ã®è¡ååŠçç 究ã«åºã¥ããŠææ ã«ããŽãªãå®çŸ©ããããã«å¯Ÿå¿ããèŠèŠçç¹åŸŽïŒç®ãè³ã錻ãéŠã®äœçœ®ãªã©ïŒãæ©æ¢°åŠç¿ã¢ãã«ã«åŠç¿ãããç¹ã§ãããŸããEquiFACSïŒéŠ¬ã®è¡šæ åäœã³ãŒãã£ã³ã°ã·ã¹ãã ïŒã®ç¥èŠã掻çšãããã詳现ãªè¡šæ 解æãå¯èœã«ããŠããŸãã4. ã©ããã£ãŠæå¹ã ãšæ€èšŒããïŒ
ã·ã¹ãã ã®æå¹æ§ã¯ä»¥äžã®æ¹æ³ã§æ€èšŒãããŸããïŒ
- ããŒã¿ã»ããïŒ440æã®éŠ¬ã®ç»åãåéãã4ã€ã®ææ ã«ããŽãªã«æåã§ã©ãã«ä»ãã
- æ€åºåšã®ç²ŸåºŠïŒ40æã®ãã¹ãç»åã§100%ã®ç²ŸåºŠãéæã
- ææ äºæž¬ã¢ãã«ã®è©äŸ¡ïŒ
- 5åå²äº€å·®æ€èšŒã§å¹³å80%ã®ç²ŸåºŠãéæã
- æçµçãªãã¹ãã»ããïŒ40æïŒã§65%ã®ç²ŸåºŠãéæã
- æ··åè¡åã«ãã詳现ãªæ§èœåæã
- GradCAMãçšããã¢ãã«ã®å€ææ ¹æ ã®å¯èŠåã
ãããã®çµæã¯ãã·ã¹ãã ã銬ã®ææ ã人éã®å°é家ã«è¿ã粟床ã§èªèã§ããããšã瀺ããŠããŸãããã ããããŒã¿ã»ããã®èŠæš¡ãéãããŠããããããããªãæ¹åã®äœå°ããããŸãã5. è°è«ã¯ããïŒ
èè ãã¯ä»¥äžã®ç¹ã«ã€ããŠè°è«ããŠããŸãïŒ
- ã·ã¹ãã ã®éçïŒããŒã¿ã»ããã®èŠæš¡ãå°ããïŒ440æïŒããšãé ãšéŠã®ã¿ãåæ察象ãšããŠããããšãªã©ã
- æ¹åã®å¯èœæ§ïŒããå€ãã®ããŒã¿åéãå šèº«ã®åæãä»ã®ã»ã³ãµãŒããŒã¿ïŒå¿ææ°ãäœæž©ãªã©ïŒã®çµ±åã
- ææ èªèã®è€éãïŒäººéã§ãé£ãã課é¡ã§ãããåç©ã§ã¯ããã«å°é£ã
- å¿çšå¯èœæ§ïŒçç£æ¥ãç£å»åŠåéã§ã®åç©çŠç¥ã®åäžã
- å«ççé æ ®ïŒåç©ã®ææ ãæ£ç¢ºã«ç解ããããšã®éèŠæ§ã
- ä»åŸã®ç 究æ¹åïŒåžæ°ç§åŠãããžã§ã¯ãã«ããããŒã¿åéããã詳现ãªææ åé¡ïŒçã¿ã®è¡šæ ãªã©ïŒã®çµ±åã
èè ãã¯ããã®ã·ã¹ãã ãåç©ã®ææ ç 究ãçŠç¥åäžã®ããã®åŒ·åãªåºç€ã«ãªããšçµè«ã¥ããŠããŸãã6. 次ã«èªãã¹ãè«æã¯ïŒ
ãã®ç 究ãããã«æ·±ããããã«ã以äžã®è«æãèªãããšããå§ãããŸãïŒ
- Wathan et al. (2015) "EquiFACS: The Equine Facial Action Coding System" - 銬ã®è¡šæ 解æã·ã¹ãã ã«ã€ããŠã®è©³çŽ°ãªç 究ã
- Gleerup et al. (2015) "An Equine Pain Face" - 銬ã®çã¿ã«é¢é£ããè¡šæ ã«ã€ããŠã®ç 究ã
- Hall et al. (2018) "Assessing Equine Emotional State" - 銬ã®ææ ç¶æ è©äŸ¡ã«é¢ããå æ¬çãªã¬ãã¥ãŒã
- Boissy et al. (2007) "Assessment of Positive Emotions in Animals to Improve Their Welfare" - åç©ã®æ£ã®ææ è©äŸ¡ã«é¢ããç 究ã
- Ren et al. (2015) "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks" - ç©äœæ€åºã¢ã«ãŽãªãºã ã®è©³çŽ°ã
ãããã®è«æã¯ã銬ã®ææ èªèã·ã¹ãã ã®åºç€ãšãªãçè«ãæè¡ã«ã€ããŠæ·±ãæŽå¯ãæäŸããŸãã
ã ãã¶ããŒã¿ãå°ãªããªããéŠããäžã ãã¿ããã ãããŸãè³ã ãã§ãããªãããããããã®ã ãã©ãåæ»ããšãããããããããšãããã®ã§ããã®ãžããå«ãããšãã粟床ãäžãããããã§ã銬ã®ææ ããããã£ãŠã®ã¯å€¢ãããã
ãšããããããã£ãŠã¿ãŠã»ãããªãšæãã®ã¯ãŽãŒã«ãã·ãããšããªã«ãã§ãŒãã«ãšããã£ãŠãäž¡æ¹ãšãã¹ãã€ãŽãŒã«ãã ãªããæ°æ§é£ãªéŠ¬ã¯ããããã¬ãŒã¹ã«ãããéäºå¿ã«ã€ãªããããšãããã°ãã«ãªã«ãªããããŠæ¶èããããšãããããããã銬ã®ã³ã³ãããŒã«ã«åœ¹ç«ã€ãšé¢çœããã§ããã
Sharing Pain: Using Pain Domain Transfer for Video Recognition of Low Grade Orthopedic Pain in Horses
ã©ããªãã®ïŒ
ãã®ç 究ã¯ã銬ã®æŽåœ¢å€ç§ççŒçãèªåçã«èªèããããã®æ©æ¢°åŠç¿ææ³ãææ¡ããŠããŸããæŽåœ¢å€ç§ççŒçã¯éŠ¬ã®å®æ¥œæ»ã®äž»èŠãªåå ã®1ã€ã§ãããæ©æçºèŠãé£ãããšãã課é¡ããããŸããç 究è ãã¯ãæ¥æ§å®éšççŒçã®ããŒã¿ã»ããããåŠç¿ãã深局åŠç¿ã¢ãã«ãçšããŠãæŽåœ¢å€ç§ççŒçããŒã¿ã»ãããžã®è»¢ç§»åŠç¿ãè¡ããŸããããã®ææ³ã«ãããç£å»åŠã®å°é家ãããé«ã粟床ã§æŽåœ¢å€ç§ççŒçãæ€åºããããšã«æåããŸããããŸãã人éã®å°é家ãšæ©æ¢°åŠç¿ã¢ãã«ã®çŒçèªèã®éãã«ã€ããŠãåæãè¡ã£ãŠããŸãã
å è¡ç 究ãšæ¯ã¹ãŠã©ãããããïŒ
æ¬ç 究ã®æ°èŠæ§ã¯ä»¥äžã®ç¹ã«ãããŸã:
- åç©ã®ç°ãªãçš®é¡ã®çŒçéã§ã®ãã¡ã€ã³è»¢ç§»ãåããŠèª¿æ»ãã
- 2ã€ã®å®äžçããŒã¿ã»ãããçšããåºç¯ãªå®éšçµæãæ瀺ããã¯ãªãŒã³ãªå®éšçšããŒã¿ã»ããããå®äžçã®è€éãªããŒã¿ãžã®ç§»è¡æã®èª²é¡ãæµ®ã圫ãã«ãã
- äžè¬çãªå€§èŠæš¡ã¢ã¯ã·ã§ã³èªèããŒã¿ã»ãããã転移ããæšæºçãªç¹åŸŽãšã銬ã®çŒçããŒã¿ã»ãããã転移ããç¹åŸŽã®æ¯èŒãè¡ã£ã
- 25æ¬ã®ãããªã¯ãªããã«å¯Ÿãã説æå¯èœæ§ã®ç 究ãè¡ãã27人ã®ç£å»åŠå°é家ã«ããããŒã¹ã©ã€ã³ãšæ©æ¢°åŠç¿ã¢ãã«ã®æ¯èŒãå®æœãã
æè¡ãææ³ã®èã¯ã©ãïŒ
æ¬ç 究ã®äž»èŠãªæè¡çã¢ãããŒãã¯ä»¥äžã®éãã§ã:
- æ¥æ§å®éšççŒçããŒã¿ã»ãã(PF)ããæŽåœ¢å€ç§ççŒçããŒã¿ã»ãã(EOP(j))ãžã®ãã¡ã€ã³è»¢ç§»åŠç¿
- 2ã€ã®æ·±å±€åŠç¿ã¢ãã«ã®äœ¿çš:
- ç³ã¿èŸŒã¿LSTMäºæµã¢ãã«(C-LSTM-2)
- KineticsäºååŠç¿æžã¿I3Dã¢ãã«
- 匱æåž«ããåŠç¿ã¢ãããŒãã®æ¢çŽ¢(æåãã)
- Grad-CAMãçšããã¢ãã«ã®å€ææ ¹æ ã®å¯èŠå
- è€æ°ã€ã³ã¹ã¿ã³ã¹åŠç¿(MIL)ãã£ã«ã¿ãçšããæšè«æã®äºæž¬æ¹å
ãããã®ææ³ãçµã¿åãããããšã§ããã€ãºã®å€ãå®äžçããŒã¿ã«ãããçŒçèªèã®æ¹åãå³ã£ãŠããŸããã©ããã£ãŠæå¹ã ãšæ€èšŒããïŒ
ç 究ã®æå¹æ§ã¯ä»¥äžã®æ¹æ³ã§æ€èšŒãããŸãã:
- 亀差æ€èšŒ: PFãšEOP(j)ã®äž¡ããŒã¿ã»ããã§åå¥ã«ãããã³ããããçµã¿åãããŠå®æœ
- ãã¡ã€ã³è»¢ç§»å®éš: PFã§åŠç¿ããã¢ãã«ãEOP(j)ã§ãã¹ã
- 人éã®å°é家ãšã®ããŒã¹ã©ã€ã³æ¯èŒ: 27人ã®ç£å»åŠå°é家ã«ãã25æ¬ã®ãããªã¯ãªããã®è©äŸ¡ãšã¢ãã«ã®äºæž¬ãæ¯èŒ
- åã ã®éŠ¬ã«å¯Ÿããæ§èœåæ: 7é ã®éŠ¬ããããã«å¯Ÿããã¢ãã«ã®æ§èœãè©äŸ¡
- å¯èŠåã«ããåæ: Grad-CAMãçšããŠã¢ãã«ã®æ³šç®é åãå¯èŠåãã人éã®å°é家ã®å€æãšæ¯èŒ
ãããã®æ€èšŒã«ãããææ¡ææ³ã人éã®å°é家ãäžåãæ§èœã瀺ããç¹ã«äœã°ã¬ãŒãã®æŽåœ¢å€ç§ççŒçã®æ€åºã«æå¹ã§ããããšã瀺ãããŸãããè°è«ã¯ããïŒ
æ¬ç 究ã§ã¯ä»¥äžã®ç¹ã«ã€ããŠè°è«ããªãããŠããŸã:
- å°é家ã®æ§èœãäœãçç±: çã芳å¯æéãæèæ å ±ã®æ¬ åŠãåäœå·®ãªã©ã圱é¿
- çµæã®äžè¬åå¯èœæ§: 13é ã®éŠ¬ãšããéããããµã³ãã«ãµã€ãºã§ã®æ€èšŒã ãããã¡ã€ã³éã®å·®ç°ãèæ ®ãããšäžå®ã®äžè¬åå¯èœæ§ã瀺å
- PFãšEOP(j)ã«ãããçŒçã®çç©åŠçå·®ç°: çã¿ã®çš®é¡ãæç¶æéã匷床ã®éããçµæã«äžãã圱é¿
- çŒç匷床ã®ã©ãã«ä»ãã®é£ãã: åç©ã®äž»èŠ³ççµéšã客芳çã«è©äŸ¡ããããšã®å°é£ã
- å®äžçããŒã¿ã®ãã€ãºãšåŒ±ã©ãã«ã®æ±ãæ¹: æž æµãªããŒã¿ããã®ãã¡ã€ã³è»¢ç§»ã®æå¹æ§
- ã¢ãã«ã¢ãŒããã¯ãã£ã®éžæ: C-LSTM-2ãI3Dãããè¯å¥œãªè»¢ç§»æ§èœã瀺ããçç±
ãããã®è°è«ãéããŠãç 究ã®éçãšä»åŸã®èª²é¡ãæ確åãããŠããŸãã次ã«èªãã¹ãè«æã¯ïŒ
æ¬ç 究ã®çºå±ãšããŠã以äžã®æ¹åæ§ã瀺åãããŠããŸã:
- ããèªç¶ãªç°å¢ã§ã®éŠ¬ã®çŒçèªè: å©èå ãå±å€ã§ã®è¡å解æ
- åç©ã®ããŒãºæšå®æè¡ã®å¿çš: Cao et al. (2019)ãMathis et al. (2021)ã®ææ³ã®æŽ»çš
- ã¹ãã¬ã¹ãšçŒçã®èå¥: Andersen et al. (2021)ã®ç 究ãããŒã¹ã«ããçºå±
- çŒçã®ãªãç¶æ ã®èªè: Lundblad et al. (2021)ã®ç 究ãšã®é¢é£
- é·æçãªçŒçè¡åã®æç³»ååæ: Rashid et al. (2020)ã®ææ³ã®æ¡åŒµ
- ãã«ãã¢ãŒãã«ãªçŒçè©äŸ¡: è¡åããŒã¿ãšççåŠçããŒã¿ã®çµ±å
ãããã®è«æãèªãããšã§ã銬ã®çŒçèªèæè¡ã®ãããªãçºå±ãšå¿çšç¯å²ã®æ¡å€§ã«ã€ãªããå¯èœæ§ããããŸãã
æŽåœ¢å€ç§ççŒçã£ãŠã®ã¯ããããããšãããã
æŽåœ¢å€ç§ççŒçãšã¯
æŽåœ¢å€ç§ççŒçãšã¯ã骚ãé¢ç¯ãçèãè ±ãé垯ãªã©ã®éååšç³»ã«é¢é£ããçã¿ãæããŸãã銬ã®å Žåã以äžã®ãããªç¹åŸŽããããŸãïŒ
å®çŸ©ïŒ
ã»éªšæ Œç³»ãçèç³»ã®åé¡ã«èµ·å ããçã¿
ã»æ¥æ§ïŒçªç¶çºçïŒãŸãã¯æ ¢æ§ïŒé·æéæç¶ïŒã®äž¡æ¹ãããäž»ãªåå ïŒ
ã»é¢ç¯çïŒç¹ã«å€åœ¢æ§é¢ç¯çïŒ
ã»éªšæãäºè£
ã»è ±ãé垯ã®æå·
ã»çèã® strainïŒé床ã®åŒã£åŒµãïŒã sprainïŒæ»æ«ïŒçç¶ïŒ
ã»è·è¡ïŒè¶³ãåŒããããäžèªç¶ãªæ©ãæ¹ïŒ
ã»ç¹å®ã®å§¿å¢ãåããé¿ãã
ã»è§Šããããšçãã
ã»é¢ç¯ãçèã®è «ã
ã»é¡ã®è¡šæ ã®å€åïŒç®ãè³ã®äœçœ®ãå£ã®ç·åŒµãªã©ïŒéèŠæ§ïŒ
ã»æ©æçºèŠãé£ãããé²è¡ãããšéçåããå¯èœæ§ããã
ã»éŠ¬ã®å®æ¥œæ»ã®äž»èŠãªåå ã®äžã€
ã»é©åãªæ²»çãè¡ãã°ãå€ãã®å Žåæ¹åãå¯èœãã®è«æã§ã¯ãæŽåœ¢å€ç§ççŒçã®äžã§ãç¹ã«äœã°ã¬ãŒãïŒè»œåºŠããäžçšåºŠïŒã®çã¿ã«çŠç¹ãåœãŠãŠããŸããããã¯ãçç¶ã埮åŠã§æç¶çã«çŸããããã人éã®ç®ã ãã§ã¯æ€åºãé£ããã±ãŒã¹ã§ãã
競走銬ã®å Žåã¯åœç¶èã«ãŸã€ãããã®ãå€ãæãã«ãªãã
ã¡ãªã¿ã«ããŒã¿ã»ããã«ã€ããŠäžã®ãŸãšãã ãšããããªãã£ãã®ã ãã©ãã©ãããåç»ã¿ããã
ç 究ã§äœ¿çšãããããŒã¿ã»ãã
ãã®ç 究ã§ã¯äž»ã«2ã€ã®ããŒã¿ã»ããã䜿çšãããŠããŸãïŒ
Pain Face (PF) ããŒã¿ã»ããïŒ
ã»6é ã®å¥åº·ãªéŠ¬ã察象
ã»æ¥æ§ã®å®éšççŒçãèªçº
ã»çã¿ã®çš®é¡ïŒ
- èè¡æ§çã¿ïŒå§è¿«ã«ãã«ããïŒ
- ççæ§çã¿ïŒã«ããµã€ã·ã³ç©è³ªã«ããç®èåºæ¿ïŒ
ã»çã¿ã®æç¶æéïŒ20åé
ã»ã©ãã«ïŒãã€ããªïŒçã¿ãã/ãªãïŒ
ã»ç·ãããªæ°ïŒ60
ã»ç·ã¯ãªããæ°ïŒ8,784
ã»ç·æéïŒçŽ9æé45åEquineOrthoPain(joint) (EOP(j)) ããŒã¿ã»ããïŒ
ã»7é ã®å¥åº·ãªéŠ¬ã察象ïŒå ã¯8é ã ã1é ã¯é€å€ïŒ
ã»è»œåºŠããäžçšåºŠã®æŽåœ¢å€ç§ççŒçãèªçº
ã»çã¿ã®èªçºæ¹æ³ïŒãªãå€ç³ïŒLPSïŒãè·è ¿é¢ç¯ïŒé£ç¯ïŒã«æ³šå°
ã»çã¿ã®æç¶æéïŒæ³šå°åŸ22-52æéã®éã§èŠ³å¯
ã»ã©ãã«ïŒ
- Composite Orthopedic Pain Scale (CPS) ã䜿çš
- 3人ã®ç¬ç«ããè©äŸ¡è ã«ããã¹ã³ã¢ïŒ0-39ïŒã®å¹³å
- ãã€ããªåé¡ã®ããã«éŸå€åŠçïŒ0ãã倧ãã = çã¿ããïŒ
ã»ç·ãããªæ°ïŒ90
ã»ç·ã¯ãªããæ°ïŒ6,710
ã»ç·æéïŒçŽ8æé41åäž¡ããŒã¿ã»ããã®ç¹åŸŽïŒ
ã»éŠ¬ã®é¡ãšäžå身ãäž»ã«æ ã£ãŠãã
ã»èã®å šäœã¯æ ã£ãŠããªã
ã»ãããªããã¯2fpsã§ãã¬ãŒã ãæœåº
ã»åã¯ãªããã¯10ãã¬ãŒã ïŒ5ç§ïŒã§æ§æç 究è ãã¯ãããã®ããŒã¿ã»ãããçšããŠãPFããŒã¿ã»ããã§åŠç¿ããã¢ãã«ãEOP(j)ããŒã¿ã»ããã«é©çšãããã¡ã€ã³è»¢ç§»åŠç¿ãè¡ããŸãããããã«ãããã¯ãªãŒã³ãªå®éšççã¿ã®ããŒã¿ãããããè€éã§åŸ®åŠãªæŽåœ¢å€ç§ççã¿ã®æ€åºãžã®ç¥è転移ãè©Šã¿ãŠããŸãã
ããããã®ãå©èãç§å Žã§æŽ»çšã§ããããã«ããŠã»ãããªãã
æ å解æã¯ãããã掻çšæ¹æ³ãããããã競走銬ã®å¥åº·ãšããæå³ã§ãããããç 究ãé²ãã§æ¬²ããåé¢ãããŒã¿ã»ãããå šç¶è¶³ããªãæãã¯ãããªãJRAãããããããŒã¿ãå ¬éããŠKaggleãšãã§ã³ã³ãããã°ããã®ã«ããç 究ææã£ãŠãã¿ããã ããå¥äŒç€Ÿãšããããããªæ°ããããã©ãã
Equimetrics -- Applying HAR principles to equestrian activities
Equimetrics - HARã®åçã銬è¡æŽ»åã«å¿çš
1. ã©ããªãã®ïŒ
Equimetricsã¯ãä¹éŠ¬æŽ»åã«ããã人éãšãŠãã®åããå æ¬çã«åæããããã®ã»ã³ãµãŒã·ã¹ãã ã§ããã©ã€ããŒã®äœ(èŽäœãé éšãè ãè)ãšãŠãã®èã«10åã®æ £æ§èšæž¬ãŠããã(IMU)ã»ã³ãµãŒãèšçœ®ãã130Hzã®ãµã³ããªã³ã°ã¬ãŒãã§3次å ã®åããææããŸããåã»ã³ãµãŒããŒãã¯MPU-6050 NEMSãšåŒã°ãã3軞ãžã£ã€ãã¹ã³ãŒãã»å é床èšãããã³ESP32ãã€ã¯ãã³ã³ãããŒã©ãŒã§æ§æãããWi-Fiçµç±ã§ããŒã¿ãéä¿¡ããŸãããŸããPIXEMã«ã¡ã©ã·ã¹ãã ã«ããæ åããŒã¿ãçµ±åããOpenPoseãã¬ãŒã ã¯ãŒã¯ãçšãã詳现ãªåäœåæãå¯èœã«ããŠããŸããåŸæ¥ã®é«äŸ¡ãªã¢ãŒã·ã§ã³ãã£ããã£ãŒæè¡ã«ä»£ããããªãŒãã³ãœãŒã¹ã®ããŒããŠã§ã¢ãšãœãããŠã¢ã掻çšããäœã³ã¹ããªãœãªã¥ãŒã·ã§ã³ãšããŠææ¡ãããŠããŸãã
2. å è¡ç 究ãšæ¯ã¹ãŠã©ãããããïŒ
å è¡ç 究ã§ã¯ãäž»ã«ãŠãã®æ©æ§èªèã«çŠç¹ãåœãŠãŠãããã¹ããŒããã©ã³ãã¹ããŒããŠã©ããã䜿çšããåçŽãªåäœèªèãäžå¿ã§ãããäŸãã°ãéãã©ã€ããŒã®æéŠã«ã»ã³ãµãŒãèšçœ®ããŠæ©æ§ãåé¡ããç 究ããå°éã®ã©ãã«ä»ãããŒã¿ã§éŠ¬ã®åäœãèªèããã¢ã«ãŽãªãºã ã®éçºãªã©ãè¡ãããŠããŸãããããã«å¯ŸããŠEquimetricsã¯ãã©ã€ããŒãšãŠãã®äž¡æ¹ã«è€æ°ã®ã»ã³ãµãŒãèšçœ®ãããã®çžäºäœçšãå æ¬çã«åæã§ããç¹ãé©æ°çã§ããç¹ã«ããŠãã®åããšã©ã€ããŒã®åããåé¢ããŠåæã§ããæ©èœã¯ãåŸæ¥ã®ã·ã¹ãã ã«ã¯ãªãç¹åŸŽã§ãããŸãããªãŒãã³ãœãŒã¹åã«ãã£ãŠç 究è ããã¬ãŒããŒã容æã«ã¢ã¯ã»ã¹ã§ããç¹ã倧ããªå©ç¹ãšãªã£ãŠããŸãã
3. æè¡ãææ³ã®èã¯ã©ãïŒ
Equimetricsã®æ žãšãªãæè¡ã¯ã以äžã®3ç¹ã§ãã第äžã«ãã©ã€ããŒãšãŠãã®åããåå¥ã«åæããææ³ã§ããç¹ã«ã©ã€ããŒã®è °ã«èšçœ®ããã»ã³ãµãŒããŒã¿ãåºæºãšããŠããŠãã®åããå·®ãåŒãããšã§ãã©ã€ããŒã®ç¬ç«ããåããæœåºããããšã«æåããŠããŸãã第äºã«ãé«ç²ŸåºŠãªè¹ã®æ¥å°ã¿ã€ãã³ã°æ€åºã§ãããŠãã®4æ¬ã®èã«èšçœ®ããã»ã³ãµãŒããåŸãããå é床èšãšãžã£ã€ãã¹ã³ãŒãã®ããŒã¿ãåå æ°è¡šçŸã«å€æããè¹ã®æ¥å°ã€ãã³ããå¹³å8.98ããªç§ã®ç²ŸåºŠã§æ€åºã§ããŸãã第äžã«ããã©ã³ã¹ãã©ãŒããŒããŒã¹ã®2ã€ã®æŽ»åèªèã¢ãã«ãå®è£ ãããŠãã®æ©æ§ãšç¹å®ã®éŠ¬å ŽéŠ¬è¡ã®åäœãé«ç²ŸåºŠã§åé¡ã§ããç¹ã§ãã
4. ã©ããã£ãŠæå¹ã ãšæ€èšŒããïŒ
ã·ã¹ãã ã®æå¹æ§ã¯ã2é ã®ãŠãã䜿çšãã45åéã®äºåããŒã¿åéã§æ€èšŒãããŸãããæ€èšŒã§ã¯ãæ©è¡ãéæ©ã襲æ©ãé害é£è¶ãªã©ã®åºæ¬çãªæ©æ§ãšåäœãå«ãæšæºåãããéŠ¬å ŽéŠ¬è¡ã®ãã¹ããããã³ã«ãå®æœããŸãããããŒã¿ã¯5ç§éã®ãŠã£ã³ããŠã§åæãããåãŠãã2åãã€ãã¹ããããã³ã«ãå®è¡ããããšã§ãã¢ãã«ã®æ±åæ§èœãåäžãããŠããŸããè©äŸ¡çµæãšããŠããŠãã®æ©æ§èªèã§ã¯F1ã¹ã³ã¢0.9324ãããè€éãªéŠ¬å ŽéŠ¬è¡ã®åäœèªèã§ã¯0.7601ãéæããŸãããç¹ã«è¹ã®æ¥å°ã¿ã€ãã³ã°ã®æ€åºç²ŸåºŠã¯ããããªããŒã¿ãšã®æ¯èŒã§å¹³å8.98ããªç§ãšããé«ç²ŸåºŠã瀺ããŸããã
5. è°è«ã¯ããïŒ
äž»ãªè°è«ç¹ãšããŠãããŒã¿ã»ããã®èŠæš¡ã«é¢ããå¶éãæããããŠããŸããçŸç¶ã§ã¯2é ã®ãŠãã®ã¿ã察象ãšããæ€èšŒã§ãããããå€æ§ãªãŠããšã©ã€ããŒã®çµã¿åããã§ã®æ€èšŒãå¿ èŠãšãããŠããŸããç¹ã«ãããŒããã¹ãšåŒã°ããæãæ¹åãžã®ç§»åãå«ãè€éãªåäœã®èªè粟床ãäœãç¹ã課é¡ãšããŠææãããŠãããããã¯ãµã³ãã«æ°ã®äžè¶³ãåå ã§ããå¯èœæ§ã瀺åãããŠããŸãããŸããé·æçãªã¹ãã«éçºãé©å¿ã®èŠ³ç¹ããã瞊æçãªç 究ã®å¿ èŠæ§ãææãããŠããŸãããããã®èª²é¡ã«å¯ŸããŠããã倧èŠæš¡ã§å€æ§ãªãµã³ãã«ã«ããæ€èšŒãšãé·æçãªé²æ远跡ã®å¿ èŠæ§ãææ¡ãããŠããŸãã
6. 次ã«èªãã¹ãè«æã¯ïŒ
- "A framework for the recognition of horse gaits through wearable devices" (Casella et al., 2020) - ãŠã§ã¢ã©ãã«ããã€ã¹ãçšãã銬ã®æ©æ§èªèã®åºç€çãã¬ãŒã ã¯ãŒã¯ã«ã€ããŠ
- "Time-Series-Based Feature Selection and Clustering for Equine Activity Recognition Using Accelerometer Data" (De Waele et al., 2023) - å é床èšããŒã¿ãçšãã銬ã®æŽ»åèªèã«é¢ããææ°ã®ç 究
- "Automatic hoof-on and -off detection in horses using hoof-mounted inertial measurement unit sensors" (Tijssen et al., 2020) - IMUã»ã³ãµãŒãçšããè¹ã®æ¥å°æ€åºã«é¢ãã詳现ãªç 究
- "Human Daily and Sport Activity Recognition Using a Wearable Inertial Sensor Network" (Hsu et al., 2018) - ãŠã§ã¢ã©ãã«ã»ã³ãµãŒãããã¯ãŒã¯ãçšããã¹ããŒã掻åèªèã®åºç€
銬ãšä¹ãæã®åäœã®é¢ä¿æ§ã枬å®ã§ãããšãããšããããã¢ã¹ãªãŒãã«ã»ã³ãµãŒä»äžããŠããããªããŒã¿ãååŸããã£ãŠã®ã¯ã以åã«æ¯ã¹ããšåºãã£ãŠããŠããããã«æããã®ã§ã競銬ã«ãããããã®ããããšãããã
ãžã§ãããŒçšã®ãã¬ãŒãã³ã°ãã·ãŒã³ãšãã«ããããã®ãåæ ããããšé¢çœããã
Statistical ranking with dynamic covariates
åçå ±å€éãçšããçµ±èšçã©ã³ãã³ã°
1. ã©ããªãã®ïŒ
ãã®è«æã¯ããã©ã±ããã»ãªã¥ãŒã¹ïŒPLïŒã¢ãã«ãšåŒã°ããçµ±èšçã©ã³ãã³ã°ã¢ãã«ãæ¡åŒµããåçãªå ±å€éïŒæéãç°å¢ãªã©ã®å€åããèŠå ïŒãçµã¿èŸŒãã æ°ããã¢ãã«ãPlusDCããææ¡ããŠããŸããPLã¢ãã«ã¯è€æ°ã®ã¢ã€ãã ãé äœä»ãããéã«äœ¿çšããã確çã¢ãã«ã§ãããåŸæ¥ã®ã¢ãã«ã§ã¯æéå€åãªã©ã®åçãªèŠå ãèæ ®ã§ããŸããã§ãããPlusDCã¢ãã«ã¯ãåã¢ã€ãã ã®åºæã®åŒ·ããè¡šãå¹çšãã¯ãã«ãšãåçãªå ±å€éã®åœ±é¿ãè¡šãä¿æ°ãã¯ãã«ãçµã¿åãããããšã§ãããæè»ãªã©ã³ãã³ã°äºæž¬ãå¯èœã«ããŸããäŸãã°ãããã¹éžæã®ã©ã³ãã³ã°ã«ãããŠãéžæã®åºæ¬çãªå®åã ãã§ãªãã幎霢ã«ããå®åã®å€åãªã©ãèæ ®ã§ããããã«ãªããŸãã
2. å è¡ç 究ãšæ¯ã¹ãŠã©ãããããïŒ
åŸæ¥ã®ç 究ã§ã¯ãïŒ1ïŒå ±å€éã®ã¿ã䜿çšããæ¹æ³ãïŒ2ïŒåºå®çãªå ±å€éãšå人å¹æãçµã¿åãããæ¹æ³ãã®2ã€ã®ã¢ãããŒããäž»æµã§ãããããããïŒ1ïŒã¯å人ã®åºæã®èœåãç¡èŠããïŒ2ïŒã¯å ±å€éãæéãšãšãã«å€åããç¶æ³ãæ±ããªããšããå¶éããããŸãããæ¬ç 究ã®PlusDCã¢ãã«ã¯ãå人ã®åºæã®èœåïŒå¹çšãã¯ãã«ïŒãšåçãªå ±å€éã®äž¡æ¹ãåæã«èæ ®ã§ããåããŠã®ã¢ãã«ã§ãããŸããçè«é¢ã§ã¯ãã¢ãã«ã®èå¥å¯èœæ§ïŒäžææ§ïŒãæå°€æšå®éã®äžèŽæ§ã«ã€ããŠãã°ã©ãçè«ãçšããæ°ããåæææ³ãææ¡ããŠããŸããããã«ãããåŸæ¥ã¯å°é£ã ã£ãå€éæ¯èŒããŒã¿ã«ãããåçèŠå ã®åœ±é¿ããçè«çã«å³å¯ã«åæããããšãå¯èœã«ãªããŸããã
3. æè¡ãææ³ã®èã¯ã©ãïŒ
æ¬ç 究ã®æè¡çãªæ žå¿ã¯ä»¥äžã®3ç¹ã§ãã第äžã«ãã¢ãã«ã®èå¥å¯èœæ§ã®æ¡ä»¶ããã°ã©ãçè«ã«ããããã«ãŒã«ã»ããªãŒãããŒãã®æŠå¿µãçšããŠç¹åŸŽä»ããŸãããããã¯ãæ¯èŒã°ã©ãã®äœçžæ§é ãšå ±å€éã®é¢ä¿ãæ確ã«ããæ°ããã¢ãããŒãã§ãã第äºã«ãæå°€æšå®éã®èšç®ã®ããã®äº€äºæ倧åã¢ã«ãŽãªãºã ãéçºããŸãããããã¯ãå¹çšãã¯ãã«ãšä¿æ°ãã¯ãã«ã亀äºã«æŽæ°ããããšã§ãèšç®å¹çãåäžãããææ³ã§ãã第äžã«ãæå°€æšå®éã®äžèŽæ§ïŒæšå®å€ãçã®å€ã«åæããããšïŒã蚌æããéã«ãã°ã©ãã®ãã§ã€ãã³ã°ææ³ãšçµéšéçšçè«ãçµã¿åãããæ°ãã蚌æææ³ãéçºããŸãããç¹ã«ãæ¯èŒã°ã©ãã®é£çµæ§ãšå ±å€éã®ééåæ§ã®æ¡ä»¶äžã§ãæšå®éã®åæé床ãæ瀺çã«ç¹åŸŽä»ããããšã«æåããŠããŸãã
4. ã©ããã£ãŠæå¹ã ãšæ€èšŒããïŒ
æ€èšŒã¯äž»ã«3ã€ã®æ¹æ³ã§è¡ãããŸããã第äžã«ãã·ãã¥ã¬ãŒã·ã§ã³å®éšã«ãã£ãŠã¢ãã«ã®çè«çæ§è³ªã確èªããŸãããå ·äœçã«ã¯ãããŒãæ°ãå¢å ããå Žåã®æå°€æšå®éã®äžèŽæ§ãåæé床ãæ€èšŒããçè«çãªäºæž¬ãšäžèŽããããšã瀺ããŸããã第äºã«ãéŠæž¯ã®ç«¶éŠ¬ããŒã¿ãçšããŠå®ããŒã¿æ€èšŒãè¡ããŸããã銬ã®ééãã¹ã¿ãŒãäœçœ®ãå ¬è¡ã®äºæ³ãšãã3ã€ã®å ±å€éãèæ ®ããåŸæ¥ã®PLã¢ãã«ãšæ¯èŒããŠäºæž¬ç²ŸåºŠãåäžããããšã瀺ããŸããã第äžã«ã1980幎ãã2024幎ãŸã§ã®ATPããã¹ã®è©ŠåããŒã¿ãåæããéžæã®å¹Žéœ¢å¹æãèæ ®ããã©ã³ãã³ã°äºæž¬ãè¡ããŸããããã®åæã§ã¯ãéžæã®ãã£ãªã¢ããŒã¯ææã®äºæž¬ãå®éã®èšé²ãšæŽåçã§ããããšã瀺ããã¢ãã«ã®å®çšæ§ãå®èšŒããŸããã
5. è°è«ã¯ããïŒ
äž»ã«2ã€ã®éèŠãªè°è«ããããŸãã1ã€ç®ã¯ãåºå®çãªå ±å€éãšåçãªå ±å€éã®éãã«ã€ããŠã®è°è«ã§ããè«æã¯ãåºå®çãªå ±å€éã¯å¹çšãã¯ãã«ã«åžåãããŠããŸããããå®è³ªçãªäºæž¬æ¹åã«ã€ãªãããªãããšãçè«çã«ç€ºããŠããŸãïŒProposition 5ïŒãããã¯ãåçãªå ±å€éãèæ ®ããããšã®å¿ èŠæ§ãè£ä»ããéèŠãªç¥èŠã§ãã2ã€ç®ã¯ãæå°€æšå®éã®åæé床ã«é¢ããè°è«ã§ããå¹çšãã¯ãã«ãšä¿æ°ãã¯ãã«ã®åæé床ãç°ãªãããšã瀺ãããããã¯2ã€ã®ãã©ã¡ãŒã¿ã®æšå®ã®é£ãããæ¬è³ªçã«ç°ãªãããšã瀺åããŠããŸãããã®ç¥èŠã¯ãå°æ¥ã®çµ±èšçæšè«ã®ç 究ã«éèŠãªç€ºåãäžããŠããŸãã
6. 次ã«èªãã¹ãè«æã¯ïŒ
以äžã®è«æ矀ãéèŠã§ãïŒ
- Fan et al. (2022) "Uncertainty quantification of MLE for entity ranking with covariates" - åºå®å ±å€éãæã€ã©ã³ãã³ã°ã¢ãã«ã®äžç¢ºå®æ§å®éåã«é¢ããåºç€çç 究
- Han and Xu (2023) "A unified analysis of likelihood-based estimators in the Plackett-Luce model" - PLã¢ãã«ã«ãããæå°€æšå®ã®çµ±äžçãªåæãæäŸ
- Singh et al. (2024) "Graphical models for cardinal paired comparisons data" - ãã¢æ¯èŒããŒã¿ã«å¯Ÿããã°ã©ãã£ã«ã«ã¢ãã«ã®ææ°ã®ç 究
ãããã®è«æã¯ãã©ã³ãã³ã°ã¢ãã«ã®çµ±èšçæ§è³ªãå ±å€éã®æ±ããã°ã©ãçè«ãšã®é¢é£æ§ã«ã€ããŠãããæ·±ãç解ãæäŸããŸãã
ããã ãã ãšã¡ãã£ãšããããããªãã£ãã®ã§æŽåã
"第äºã«ãéŠæž¯ã®ç«¶éŠ¬ããŒã¿ãçšããŠå®ããŒã¿æ€èšŒãè¡ããŸããã銬ã®ééãã¹ã¿ãŒãäœçœ®ãå
¬è¡ã®äºæ³ãšãã3ã€ã®å
±å€éãèæ
®ããåŸæ¥ã®PLã¢ãã«ãšæ¯èŒããŠäºæž¬ç²ŸåºŠãåäžããããšã瀺ããŸããã" ãšãããšããããã£ãšè©³ããã
詳现ãªå®éšçµæã®è§£èª¬
PlusDCã¢ãã«ã®å®äžçã§ã®æå¹æ§ã瀺ãããã«ã1999幎ãã2005幎ã®éŠæž¯ç«¶éŠ¬ããŒã¿ãçšãã詳现ãªåæãè¡ãããŸãããå®éšã®å ·äœçãªå 容ã¯ä»¥äžã®éãã§ãïŒ
ããŒã¿ã»ããã®ç¹åŸŽïŒ
- ç·ãµã³ãã«æ°ïŒ6,328ã¬ãŒã¹
- åæ察象ã®éŠ¬ïŒ2,814é
- åã¬ãŒã¹ã®åºèµ°é æ°ïŒ4é ãã14é ã®ç¯å²
- å銬ã®æå°åºèµ°åæ°ïŒ10åïŒå¯Ÿæ°nãã倧ããïŒ
èæ ®ããã3ã€ã®å ±å€éïŒ
- å®éã®éãïŒéŠ¬ãèè² ãééïŒïŒå¯Ÿæ°å€æåŸãäžå¿åãšæ£èŠåãå®æœ
- æ çªïŒã¬ãŒã¹ã§ã®ã¹ã¿ãŒãäœçœ®ïŒïŒåã¬ãŒã¹ã®åºèµ°é æ°ã«åºã¥ããŠ0-1ã®ç¯å²ã«æ£èŠå
- å ¬è¡äºæ³ïŒã¬ãŒã¹çŽåã®ãªããºããç®åºãããåçïŒïŒå¯Ÿæ°å€æãå®æœ
åæçµæïŒ
- æšå®ãããä¿æ°ãã¯ãã«ïŒ(-0.139, -0.188, 0.502)
- ééïŒè² ã®åœ±é¿ïŒéãã»ã©äžå©ïŒ
- æ çªïŒè² ã®åœ±é¿ïŒå€åŽã®æ ã»ã©äžå©ïŒ
- å ¬è¡äºæ³ïŒæ£ã®åœ±é¿ïŒäºæ³ãé«ãã»ã©æå©ïŒ
ã¢ãã«æ§èœè©äŸ¡ïŒ
- 63åå²äº€å·®æ€èšŒã§è©äŸ¡ãå®æœ
- è©äŸ¡ææšïŒTop-1äºæž¬ãTop-3äºæž¬ãå®å šé äœäºæž¬ããããã®å¯Ÿæ°å°€åºŠ
- æ¯èŒå¯Ÿè±¡ïŒPLã¢ãã«ãå ¬è¡äºæ³
- çµæïŒ
- ãã¹ãŠã®è©äŸ¡ææšã§PlusDCã¢ãã«ãPLã¢ãã«ãäžåã
- Top-1ãTop-3äºæž¬ã§ã¯å ¬è¡äºæ³ãšåç
- å®å šé äœäºæž¬ã§ã¯å ¬è¡äºæ³ã倧ããäžåã
ãã®å®éšçµæã¯ãPlusDCã¢ãã«ãå®äžçã®è€éãªèŠå ãå¹æçã«æããããæ£ç¢ºãªäºæž¬ãå¯èœã§ããããšã瀺ããŠããŸããç¹ã«ãã¢ãã«ãæããåèŠå ã®åœ±é¿åºŠã¯ã競銬ã®äžè¬çãªçµéšåãšãæŽåçã§ãããã¢ãã«ã®è§£éå¯èœæ§ãé«ãããšã瀺ãããŸããã
ãªãã»ã©ãæ¢åã®çµ±èšã¢ãã«ãããåªããŠããã£ãŠã®ã¯ããããã ãã©ãæ€éãæ çªã«ã€ããŠã¯ããã¯åœç¶ã ãã人æ°ãé«ãïŒçäžçã¯é«ãããªããºã¯äœããã«ãªãã®ã§ããããããã§ããããšããæ°ãããã ããããå°ã競銬ãšããã銬åžçã«ã©ãããæå³åãããããïŒ
ããã銬åžäºæ³ã«æŽ»ããã«ã¯ã©ããã颚ã«èããã°ããã§ããããïŒ
銬åžäºæ³ãžã®æŽ»çšã«é¢ããèå¯
â»ããã¯åŠè¡çãªèå¯ã§ãããå®éã®éŠ¬åžè³Œå ¥ãæšå¥šãããã®ã§ã¯ãããŸããã
- ã¢ãã«ããåŸãããéèŠãªç¥èŠ
- ééã®åœ±é¿ïŒä¿æ°:-0.139ïŒ
- æ€éãå¢ããã»ã©äžå©ã«ãªã
- ç¹ã«ãåã銬ã®åèµ°ãšã®æ€éå·®ã«æ³šç®ãå¿ èŠ
- æ çªã®åœ±é¿ïŒä¿æ°:-0.188ïŒ
- å€æ ã«ãªãã»ã©äžå©ã«ãªãåŸå
- ãã ããããã¯å šã¬ãŒã¹å¹³åã®åŸåã§ãããã³ãŒã¹ç¹æ§ã«ããå·®ç°ã¯èæ ®ãå¿ èŠ
- å ¬è¡äºæ³ã®éèŠæ§ïŒä¿æ°:0.502ïŒ
- æã圱é¿åã®å€§ããèŠå
- ãªããºã«åæ ãããæ å ±ã®ä¿¡é Œæ§ãé«ãããšã瀺å
- ã¢ãã«ã®ç¹åŸŽçãªæ§èœ
- Top-1ãTop-3äºæž¬ã§ã¯å ¬è¡äºæ³ãšåçã®æ§èœ
- ååã»è€åã®äºæ³ã«ã¯ããªããºæ å ±ã ãã§ãååãªå¯èœæ§
- å®å šé äœäºæž¬ã§åªããæ§èœ
- 銬é£ãäžé£è€ãäžé£åãªã©ã®é äœé¢ä¿ãéèŠãªéŠ¬åžã«æçšãªå¯èœæ§
- ç¹ã«ããªããºã§ã¯èŠèœãšãããããäžäœïœäžäœã®é äœäºæ³ãæ£ç¢º
- å®è·µçãªæŽ»çšã«ãããçæç¹
- ã¢ãã«ã®éçã®èªè
- 倩åãéŠ¬å Žç¶æ ã調æåž«ã®è©äŸ¡ãªã©ãå«ãŸããŠããªãèŠå ãå€æ°ååš
- ããŒã¿ã®æéçå¶çŽïŒ1999-2005幎ïŒã«ãããçŸåšã®ç«¶éŠ¬ãšã¯ç°å¢ãç°ãªãå¯èœæ§
- æåŸ å€ã®èãæ¹
- ã¢ãã«ã®äºæž¬ãšãªããºã®ä¹é¢ã倧ããå Žåã«æè³æ©äŒãããå¯èœæ§
- ãã ãããªããºèªäœãéèŠãªäºæž¬å åã§ããããšã«æ³šæ
- ããé«åºŠãªæŽ»çšæ¹æ³ã®å¯èœæ§
- åçå ±å€éã®æ¡åŒµ
- éŠ¬å Žç¶æ ãæ°æž©ã湿床ãªã©ã®ç°å¢èŠå ã®çµã¿èŸŒã¿
- 調æããŒã¿ãè¡çµ±æ å ±ãªã©ã®è¿œå
- åå¥ã³ãŒã¹ç¹æ§ã®èæ ®
- çŽç·ãèãããŒããªã©ã®ã³ãŒã¹ç¹æ§ããšã®ä¿æ°èª¿æŽ
- è·é¢å¥ã®åœ±é¿åºŠã®å€åã®åæ
ãã®ã¢ãã«ã¯åŠè¡ç 究ç®çã§éçºãããŠãããå®éã®éŠ¬åžè³Œå ¥ãžã®å¿çšã«ã¯ãããå€ãã®å®åçãªæ€èšŒãšè¿œå ã®åæãå¿ èŠã§ãããŸããã®ã£ã³ãã«ã«ã¯åžžã«ãªã¹ã¯ã䌎ãããšãèªèããããšãéèŠã§ãã
ããã ããã競銬ã®å Žåã¯ä»ã«ãå€åããèŠçŽ ãå€ãããæåŸ å€ã£ãŠããèãæ¹ã¯ããã«ã¯å«ãŸããŠããªãã®ã§ãããã§éŠ¬åžã©ããããšãã話ã§ã¯ãªãããã
- å®å šé äœäºæž¬ã§åªããæ§èœ
- 銬é£ãäžé£è€ãäžé£åãªã©ã®é äœé¢ä¿ãéèŠãªéŠ¬åžã«æçšãªå¯èœæ§
- ç¹ã«ããªããºã§ã¯èŠèœãšãããããäžäœïœäžäœã®é äœäºæ³ãæ£ç¢º
ãšãããšããã¯é¢çœããããã次走ã®äºæ³ã«æŽ»ããããããããªãã
A generative approach to frame-level multi-competitor races
ãã¬ãŒã ã¬ãã«ã®ãã«ãã³ã³ããã£ã¿ãŒã¬ãŒã¹ã«å¯Ÿããçæçã¢ãããŒã
1. ã©ããªãã®ïŒ
ãã«ãã³ã³ããã£ã¿ãŒ(è€æ°ç«¶äºè )ã¬ãŒã¹ã«ãããŠãã¬ãŒã¹å ã®æŠç¥ãèæ ®ããå æ¬çãªçæã¢ãã«ãææ¡ããç 究ã§ããåŸæ¥ã®ç 究ã§ã¯ã¬ãŒã¹çµæã®ã¿ãæ±ã£ãŠããŸããããæ¬ç 究ã§ã¯ãã¬ãŒã ã¬ãã«(1ç§ããã1-25ãã¬ãŒã )ã®è©³çŽ°ãªãã©ããã³ã°ããŒã¿ã掻çšãã競äºè ã®äœçœ®é¢ä¿ãæŠç¥çãªåããã¢ãã«åããŠããŸããç¹ã«éŠ¬ã®åé²è·é¢ãšæšªæ¹åã®åããåããŠèããããšã§ããã©ããã£ã³ã°(颚ããå¹æ)ãªã©ã®æŠç¥çèŠçŽ ãåãå ¥ããŠããŸãããã®ã¢ãã«ã«ãããå®éã®ã¬ãŒã¹ããŒã¿ããã®äºæž¬ã ãã§ãªããä»®æ³çãªã¬ãŒã¹ã·ããªãªã®ã·ãã¥ã¬ãŒã·ã§ã³ãå¯èœã«ãªããŸããããã¥ãŒãšãŒã¯ç«¶éŠ¬åäŒ(NYRA)ãæäŸãã2019幎ã·ãŒãºã³ã®1ãã€ã«ã¬ãŒã¹ã®ããŒã¿ãçšããŠæ€èšŒããŠããŸãã
2. å è¡ç 究ãšæ¯ã¹ãŠã©ãããããïŒ
åŸæ¥ã®ç 究ã¯äž»ã«é äœããŒã¿ã®ã¿ãæ±ããã¬ãŒã¹å ã®æŠç¥çèŠçŽ ãèæ ®ã§ããŠããŸããã§ãã[18,19,23,28]ããŸãã競äºè ã®èœåè©äŸ¡ãæç³»åã§å€åããç¶æ 空éã¢ãã«ãšããŠæ±ãããŠããŸãã[11,14-17]ãæ¬ç 究ã®é©æ°çãªç¹ã¯ããã¬ãŒã ã¬ãã«ã®ãã©ããã³ã°ããŒã¿ã掻çšããŠç«¶äºè éã®çžäºäœçšãæŠç¥ãæ瀺çã«ã¢ãã«åããç¹ã§ããç¹ã«ãåé²ãšæšªæ¹åã®åããåé¢ããŠã¢ãã«åããããšã§ããã©ããã£ã³ã°ãå åŽã¬ãŒã³ã®æå©ããªã©ã®æŠç¥çèŠçŽ ãçµã¿èŸŒãããšãã§ããŸããããŸããå®å šãªçæã¢ãã«ãšããããšã§ãä»®æ³çãªã¬ãŒã¹ã·ããªãªã®ã·ãã¥ã¬ãŒã·ã§ã³ãå¯èœã«ãªããæ§ã ãªæŠç¥ã®å¹æãå®éçã«è©äŸ¡ã§ããããã«ãªããŸããã
3. æè¡ãææ³ã®èã¯ã©ãïŒ
æ¬ç 究ã®æ žãšãªãæè¡ã¯ã以äžã®3ç¹ã§ãã1ã€ç®ã¯ã競äºè ã®åããåé²è·é¢ãšæšªæ¹åã®ç§»åã«åé¢ããŠã¢ãã«åããææ³ã§ããããã«ãããã¬ãŒã³åãããã©ããã£ã³ã°ãªã©ã®æŠç¥çèŠçŽ ãè¡šçŸã§ããŸãã2ã€ç®ã¯ã競äºè ã®èœåãã¬ãŒã¹ã®åãã§ãŒãºã§æ»ããã«å€åããBã¹ãã©ã€ã³é¢æ°ã§ã¢ãã«åããŠããç¹ã§ããããã«ãããã¹ã¿ãŒãçŽåŸã®å éãçµç€ã®ã¹ããŒããªã©ãã¬ãŒã¹ã®ç°ãªãå Žé¢ã§ã®èœåãè¡šçŸã§ããŸãã3ã€ç®ã¯ãOpenFOAMãçšããæµäœååŠã·ãã¥ã¬ãŒã·ã§ã³ã«ããããã©ããã£ã³ã°å¹æãå®éåããŠããç¹ã§ãããããã®èŠçŽ ããã€ãºã¢ãã«ãšããŠçµ±åããStan[5]ãçšããŠå¹ççã«å®è£ ããŠããŸãã
4. ã©ããã£ãŠæå¹ã ãšæ€èšŒããïŒ
ã¢ãã«ã®è©äŸ¡ã¯3ã€ã®ã¬ãã«ã§è¡ãããŠããŸãã1ã€ç®ã¯ããã¬ãŒã ã¬ãã«ã§ã®äºæž¬ç²ŸåºŠã®è©äŸ¡ã§ãã2ã€ç®ã¯ãã¬ãŒã¹å šäœã®çµæ(é äœãå®èµ°æé)ã®äºæž¬ç²ŸåºŠã®è©äŸ¡ã§ãã3ã€ç®ã¯ãæšå®ããããã©ã¡ãŒã¿ãçæãããã·ãã¥ã¬ãŒã·ã§ã³ã®åŠ¥åœæ§ã®è©äŸ¡ã§ããå ·äœçãªæ€èšŒãšããŠãåçãªåçäºæž¬ã®ç²ŸåºŠããžã§ãããŒã®è©äŸ¡ã©ã³ãã³ã°ã®åŠ¥åœæ§(å®éã®æ瞟ãšã®æ¯èŒ)ã銬ã®ãããã¡ã€ã«åé¡ã®æå³åããã¹ã¿ãŒãã¬ãŒã³ã®åœ±é¿åæãªã©ãè¡ãããŠããŸããç¹ã«ãIrad Ortiz Jr.ãªã©äžäœãžã§ãããŒã®è©äŸ¡ãå®éã®æ瞟ãšæŽåçã§ããããšããã¹ã¿ãŒãã¬ãŒã³ã®åœ±é¿åæãçŽæçãªç解ãšäžèŽããããšã瀺ãããŠããŸãã
5. è°è«ã¯ããïŒ
äž»ãªè°è«ãšããŠãã¢ãã«ã®èšç®å¹çãšé©çšå¯èœæ§ã«é¢ãã課é¡ãæããããŠããŸãã1ãã€ã«ã¬ãŒã¹ã®å ŽåãçŽ400ãã¬ãŒã ã®ããŒã¿ãããã6é ã®éŠ¬ã§2000åã®ã·ãã¥ã¬ãŒã·ã§ã³ãè¡ããš960äžåã®èšç®ãå¿ èŠã«ãªããŸãããã®èšç®éã®èª²é¡ã«å¯ŸããŠãStanã®C++å®è£ ã掻çšããããšã§ã90-120ç§çšåºŠãŸã§èšç®æéãççž®ã§ãããšå ±åããŠããŸãããŸããããé·æéã®ã¬ãŒã¹(èªè»¢è»ããŒãã¬ãŒã¹ãªã©)ãžã®é©çšã«åããŠããµã³ããªã³ã°ã¹ããŒã ããšãã¥ã¬ãŒã¿ã®æŽ»çšãææ¡ãããŠããŸããããã«ããã©ããã£ã³ã°ã¢ãã«ã®ç°¡ç¥å(颚åãã®åçŽåãªã©)ã«ã€ããŠãè°è«ãããŠããããã粟緻ãªã¢ãã«åã®äœå°ã瀺åãããŠããŸãã
6. 次ã«èªãã¹ãè«æã¯ïŒ
æ¬ç 究ã®çºå±ãšããŠã以äžã®è«æãåèã«ãªããŸãïŒ
- ãã¹ã±ããããŒã«ã«ãããæåŸ ææ䟡å€(EPV)ã®èãæ¹ïŒCervone et al. (2016) [6]
- ãã«ãã³ã³ããã£ã¿ãŒã²ãŒã ã«ãããéžæè©äŸ¡ïŒChe and Glickman (2022) [7]
- åçç¶æ 空éã¢ãã«ã«ãã競æè è©äŸ¡ïŒGlickman (1999) [14]
- 銬ã®ãã©ããã£ã³ã°å¹æã®ç©ççç 究ïŒSpence et al. (2012) [29]
- ã·ãã¥ã¬ãŒã·ã§ã³ã®ãšãã¥ã¬ãŒã·ã§ã³ã«é¢ããç 究ïŒThiagarajan et al. (2020) [31]
ãããããé¢çœããïŒã·ãã¥ã¬ãŒã·ã§ã³ãçæããã£ãŠããšã ããã
èªåã®ç«¶éŠ¬äºæ³ã§ã¯ãã©ã®ãããªéåã«ãªããïŒãéèŠã ãšæã£ãŠããŠãã¹ã¿ãŒãããã·ã¥ã®éããšãæ é ãšãããŸãèããŠãããããéŠ¬å Žå«ããŠããŒã¹ãã©ããªããïŒã¿ãããªæãã§èããã
ãã ãåœç¶ãã®ãããã¯ãžã§ãããŒãèªèããŠãããšæãã®ã§ããžã§ãããŒã®æŠç¥ãã¬ãŒã¹ã«ã©ããã圱é¿ãã§ããïŒã¿ãããªäºæ³ã¯ç¡çã ããªããšæã£ãŠããŠãæŠããžã§ãããŒã®ç¹åŸŽã¿ãããªãã®ãéã«èæ ®ãããããããããªããäŸãã°ã岩ç°åº·Jãå ãçªãã®ãäžæããäž¹å Jã¯ãããã¯ãããšãããçµ±èšçãªåŸåã¯ããã®ãããããªããã©ããããã©ãããã圱é¿ããã®ããšãããããªããããã«ã¡ãŒã«JããŸããããšãã¯ããããã ãã©ã
ãã®ãžãã®ããžã§ãããŒããšã®æŠç¥ãšããã·ãã¥ã¬ãŒã·ã§ã³ã§èŠããããã«ãªããšãããç¹åŸŽãéç«ã€ããã«ãªããããããªããã