Best 11 selection using machine learning

Abstract

Cricket has become the most popular game not only in Bangladesh but also around the world. Day by day it is gaining more people’s attention. The name and glory of team Bangladesh has trespassed the country’s border and it is creating a huge impact in the world cricket. To make sure the continuous development of the team, a more professional approach and help from IT sector is needed. Remembering this fact, we used machine learning to select the best players from the standby list based on the previous playing statistics and then from that players we have found out the winning team combination. We have collected the data from websites that offers trustable sports statistics. Feature selection algorithms like Recursive feature elimination and univariate selection are used to find out the attributes that are more related to the output feature. Machine learning Algorithms such as linear regression, support vector machine with linear and polynomial kernel was used to predict the runs scored by a batsman and runs given by a bowler. Later on, we have also used fully connected neural network to find out the performance comparison of different algorithm. We have selected the players for the team according to their performances and experiences. Our goal was to form a well-balanced team through our approach.