论文信息 - Robust Full Bayesian Methods for Neural Networks

Robust Full Bayesian Methods for Neural Networks

In this paper, we propose a full Bayesian model for neural networks. This model treats the model dimension (number of neurons), model parameters, regularisation parameters and noise parameters as random variables that need to be estimated. We then propose a reversible jump Markov chain Monte Carlo (MCMC) method to perform the necessary computations. We find that the results are not only better than the previously reported ones, but also appear to be robust with respect to the prior specification. Moreover, we present a geometric convergence theorem for the algorithm.

[1] Wray L. Buntine,et al. Bayesian Back-Propagation , 1991, Complex Syst..

[2] David J. C. MacKay,et al. A Practical Bayesian Framework for Backpropagation Networks , 1992, Neural Computation.

[3] L. Tierney. Markov Chains for Exploring Posterior Distributions , 1994 .

[4] J. Besag,et al. Bayesian Computation and Stochastic Systems , 1995 .

[5] P. Green. Reversible jump Markov chain Monte Carlo computation and Bayesian model determination , 1995 .

[6] Geoffrey E. Hinton,et al. Bayesian Learning for Neural Networks , 1995 .

[7] C. C. Homes,et al. Bayesian Radial Basis Functions of Variable Dimension , 1998, Neural Computation.

[8] Peter Müller,et al. Feedforward Neural Networks for Nonparametric Regression , 1998 .

[9] Christophe Andrieu,et al. Robust Full Bayesian Learning for Neural Networks , 1999 .

[10] Arnaud Doucet,et al. Sequential Monte Carlo Methods to Train Neural Network Models , 2000, Neural Computation.