Stacking-based deep neural network for player scouting in football 1

In [8], the authors propose to Datascouting is one of the most known data evaluate individually the players according to their data applications in professional sport, and specifically and then to estimate their integration in a given team football. Its objective is to analyze huge database of with Random Forest algorithm. Even if we also use players in order to detect high potentials that can be supervised machine learning algorithm, our approach is then individually considered by human scouts. In this very different because our automatic labelling process paper, we propose a stacking-based deep learning is based on the evolution of the player in the time and model to detect high potential football players. Applied the algorithms we used are stacked deep neural on open-source database, our model obtains networks which allows accurate identifications of the significantly better results that classical statistical most promising players. In Section II, we describe the methods.