We propose a new method for comparing learning algorithms on multiple tasks which is based on a novel non-parametric test that we call the Poisson binomial test. The key aspect of...
We propose a novel type of document classification task that quantifies how much a given document (review) appreciates the target object using not binary polarity (good or bad) b...
We consider two software packages that interact with each other as components: Overture and PETSc. An interface between these two packages could be of tremendous value to applicati...
Kristopher R. Buschelman, William Gropp, Lois C. M...
The recently introduced online confidence-weighted (CW) learning algorithm for binary classification performs well on many binary NLP tasks. However, for multi-class problems CW l...
There are well-established methods for reducing the number of support vectors in a trained binary support vector machine, often with minimal impact on accuracy. We show how reduce...