Andrej Panjan (2009) Prediction of Successfulness of Tenis Players Using Machine Learning. EngD thesis.
In this thesis, a survey of prediction of successfulness of Slovenian tennis players using machine learning methods is conducted. Tests are performed for three different selections, in male and female competition. Machine learning is becoming an increasingly important tool for knowledge discovery in data. The reason for this is a rapid increase of data that is collected in recent years. Since more and more measurements are performed in tennis in recent years, we decided to do a survey of prediction of successfulness of Slovenian tennis players. The results obtained will be of assistance to tennis profession in the education of young tennis players. At the beginning a problem of prediction of tennis player successfulness is presented. Factors that affect the successfulness are presented briefly, while the factors that are used in the study are discussed in detail. The study focuses on the motor and morphological factors. The objectives of this survey, the machine learning theory used in the solution and working methodology used in the survey are presented next. The results and the interpretation of prediction of successfulness of tennis players for current and future age periods are given at the end. For prediction, classification and regression methods of machine learning with two different approaches for optimal attribute subset selection were used.
Actions (login required)