Modeling Unsupervised Learning with SUSTAIN
Todd Matthew Gureckis, Bradley C. Love · 2002
SUSTAIN (Supervised and Unsupervised STratified Adaptive Incremental Network) is a network model of human category learning. This paper extends SUSTAIN so that it can be used to model unsupervised learning data. A modified recruitment mechanism is introduced that creates new conceptual clusters in response to sur-prising events during learning. Two seemingly contra-dictory unsupervised learning data sets are modeled us-ing this new recruitment method. In addition, the fea-sibility of using a unified recruitment method for both supervised and unsupervised learning is discussed.