Bowler performance prediction for one-day international cricket using neural networks

Selecting the right bowlers against the right team plays an important role in one-day international cricket tournaments. Accurate prediction of the runs given and wickets taken by a bowler plays a potent role in the bowler selection process. A neural network approach using backpropagation network (BPN) and radial basis function network (RBFN) was explored to predict the performance of the Indian cricket team bowlers. Prediction models worked more effectively for runs than for wickets. Consequently, a classification approach was designed for the wickets scenario using the two network paradigms specified above. Performance of BPN and RBFN models was compared for the prediction and classification scenario.