Using Support Vector Machines to Determine Linear Separability

Abstract

Most theories on categorization agree on a main principle: category learning involves creating a category representation, and categorizing an item involves comparing that item to the representations of different categories. The theories, however, disagree on the nature of these category representations. There are two main competing lines of thought on category representations: exemplar-based theories and prototype-based theories (Valentine, 1991).


Back to Table of Contents