Content deleted Content added
Line 61:
:<math>f(x)=\sum^n_{i=0}\alpha_i\sum^p_{m=1}\eta_mK_m(x_i^m,x^m)</math>
<math>\eta</math> can be modeled with a Dirichlet prior and <math>\alpha</math> can be modeled with a zero-mean Gaussian and an inverse gamma variance prior. This model is then optimized using a customized [[multinomial probit]] approach using a [[Gibbs sampling|Gibbs sampler]].
<ref>Mark Girolami and Simon Rogers. Hierarchic Bayesian models for kernel learning. In Proceedings of the 22nd International Conference on Machine Learning, 2005</ref>.
These methods have been used successfully in applications such as protein fold recognition and protein homology problems <ref>Theodoros Damoulas and Mark A. Girolami. Combining feature spaces for classification. Pattern
| |||