Table tennis video data mining based on performance optimization of artificial fish swarm algorithm
Hai Wang 1, Wenbo Qu 2 , Qunli Shen3
COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(12A) 584-588
1 Department of Physical Education, Shanghai Business School, Shanghai, 201400, China
2 School of Humanities and Law, Shanghai Business School, Shanghai, 201400, China
3 College of Computer and Information , Shanghai Business School, Shanghai, 201400, China
In view of the traditional AFSA still has problems such as the low optimal performance and poor efficiency in data mining for table tennis video, a video data mining model for table tennis match is proposed based on improved AFSA. First, the traverse range of chaotic motion enlarges to the value range of AFSA optimization variables by leading in chaotic Logistic mapping. And then it increases the optimal artificial fish state information on the basis of original artificial fish behaviour, so guides the artificial fish to quickly close to the global optimal, improves the speed of the algorithm. Finally, adaptive optimize the search strategy of traditional AFSA, and apply the improved algorithm to the video data mining for table tennis match. The simulation experiments show that the video data mining model for table tennis match based on improved AFSA has better ability of optimization, and can dig out more attributes and types in the table tennis game video.