Work

Preventing Over-Fitting during Model Selection via Bayesian Regularisation of the Hyper-Parameters