Accurate Bayesian Data Classification without Hyperparameter Crossvalidation
This topic contains 0 replies, has 1 voice, and was last updated by arXiv 1 year ago.

Accurate Bayesian Data Classification without Hyperparameter Crossvalidation
We extend the standard Bayesian multivariate Gaussian generative data classifier by considering a generalization of the conjugate, normalWishart prior distribution and by deriving the hyperparameters analytically via evidence maximization. The behaviour of the optimal hyperparameters is explored in the highdimensional data regime. The classification accuracy of the resulting generalized model is competitive with stateofthe art Bayesian discriminant analysis methods, but without the usual computational burden of crossvalidation.
Accurate Bayesian Data Classification without Hyperparameter Crossvalidation
by M Sheikh, A C C Coolen
https://arxiv.org/pdf/1712.09813v1.pdf
You must be logged in to reply to this topic.