Analytic Solution of Hierarchical Variational Bayes in Linear Inverse Problem
In singular models, the Bayes estimation, commonly, has the advantage of the generalization performance over the maximum likelihood estimation, however, its accurate approximation using Markov chain Monte Carlo methods requires huge computational costs. The variational Bayes (VB) approach, a tractable alternative, has recently shown good performance in the automatic relevance determination model (ARD), a kind of hierarchical Bayesian learning, in brain current estimation from magnetoencephalography (MEG) data, an ill-posed linear inverse problem. On the other hand, it has been proved that, in three-layer linear neural networks (LNNs), the VB approach is asymptotically equivalent to a positive-part James-Stein type shrinkage estimation. In this paper, noting the similarity between the ARD in a linear problem and an LNN, we analyze a simplified version of the VB approach in the ARD. We discuss its relation to the shrinkage estimation and how ill-posedness affects learning. We also propose the algorithm that requires simpler computation than, and will provide similar performance to, the VB approach.
KeywordsLinear Inverse Problem Shrinkage Estimation Huge Computational Cost Automatic Relevance Determination Hierarchical Variational
Unable to display preview. Download preview PDF.
- 1.Hinton, G.E., van Camp, D.: Keeping Neural Networks Simple by Minimizing the Description Length of the Weights. In: Proc. of COLT, pp. 5–13 (1993)Google Scholar
- 2.Attias, H.: Inferring Parameters and Structure of Latent Variable Models by Variational Bayes. In: Proc. of UAI (1999)Google Scholar
- 4.Sato, M., Yoshioka, T., Kajihara, S., Toyama, K., Goda, N., Doya, K., Kawato, M.: Hierarchical Bayesian Estimation for MEG inverse problem. Neuro Image 23, 806–826 (2004)Google Scholar
- 5.James, W., Stein, C.: Estimation with Quadratic Loss. In: Proc. of the 4th Berkeley Symp. on Math. Stat. and Prob., pp. 361–379 (1961)Google Scholar
- 6.Nakajima, S., Watanabe, S.: Generalization Error and Free Energy of Variational Bayes Approach of Linear Neural Networks. In: Proc. of ICONIP, Taipei, Taiwan, pp. 55–60 (2005)Google Scholar
- 9.Nakajima, S., Watanabe, S.: Generalization Performance of Subspace Bayes Approach in Linear Neural Networks. IEICE Trans. E89-D, 1128–1138 (2006)Google Scholar