Interpolation Models with Multiple Hyperparameters
- 287 Downloads
A traditional interpolation model is characterized by the choice of regularizer applied to the interpolant, and the choice of noise model. Typically, the regularizer has a single regularization constant α, and the noise model has a single parameter β. The ratio α/β alone is responsible for determining globally all these attributes of the interpolant: its ‘complexity’, ‘flexibility’, ‘smoothness’, ‘characteristic scale length’, and ‘characteristic amplitude’. We suggest that interpolation models should be able to capture more than just one flavour of simplicity and complexity. We describe Bayesian models in which the interpolant has a smoothness that varies spatially. We emphasize the importance, in practical implementation, of the concept of ‘conditional convexity’ when designing models with many hyperparameters.
KeywordsTraditional Model Noise Model Markov Chain Monte Carlo Method Artificial Data Characteristic Amplitude
Unable to display preview. Download preview PDF.
- Blake, A., and Zisserman, A. (1987) Visual Reconstruction. Cambridge Mass.: MIT Press.Google Scholar
- Gull, S. F. (1988) Bayesian inductive inference and maximum entropy. In Maximum Entropy and Bayesian Methods in Science and Engineering, vol. 1: Foundations, ed. Gull, S. F, pp. 53 – 74, Dordrecht. Kluwer.Google Scholar
- MacKay, D. J. C. (1994) Hyperparameters: Optimize, or integrate out? In Maximum Entropy and Bayesian Methods, Santa Barbara 1993, ed. by G. Heidbreder, Dordrecht. Kluwer.Google Scholar
- MacKay, D. J. C, and Takeuchi, R., (1994) Interpolation models with multiple hyperparameters. Submitted to IEEE PAMI.Google Scholar