
Recent Posts
Recent Comments
Archives
Categories
Meta
Monthly Archives: May 2015
Modeling atop a document representation
The paper “DiscLDA: Discriminative Learning for Dimensionality Reduction and Classification” [1] describes a model that not only generates documents but learns them by associating each document with a label. The discrimination of a document is a function of the generative … Continue reading
Posted in optimization, statistics
Tagged discriminative, documents, generative, modeling, statistics
Leave a comment
Entropic Priors
Dirichlet (either by itself, or as a mixture of, or as a hierarchy of) priors are by no means the only option of controlling sparsity of topic mixtures. Entropic priors stand out as an interesting alternative. Given a probability distribution … Continue reading
Optimizing the Dirichlet hyperparameters
One of the things you’ll notice in papers describing generative models of documents using a Dirichlet prior is to simply fix the Dirichlet hyperparameter that controls the distributions of topic mixtures for each document. This isn’t ideal when you wish … Continue reading
Posted in optimization, statistics
Tagged dirichlet, hyperparameter, optimization, statistics
Leave a comment