Using video analysis and artificial neural network to explore association rules and influence scenarios in elite table tennis matches

To become an elite table tennis player, aside from continually practicing, players must know their strengths and weaknesses to plan their strategy beforehand and increase their winning rate. The main problems with previous research were that the data collected were incomplete and imprecise. To address these problems, we established "The Intellectual Tactical System in Competitive Table Tennis", using video analysis to collect competitive data. Additionally, we proposed a machine learning method using a combination of feature-selection and association rules to discover interesting rules from the data. The international matches of the Taiwanese table tennis single player Yun-Ju Lin were used as research samples by applying 3 S (speed, spin, spot) theory to collect and analyze data. The critical factors and scenarios were analyzed to identify the winning tactical models. The results of this study may provide useful suggestions for Yun-Ju Lin on training and building tactics in competitions. The similar approach may be essential for elite players and coaches to have appropriate tactical analysis.