Text this: Comparison of Bayesian learning and conjugate gradient descent training of neural networks