Environment and Goals Jointly Direct Category Acquisition

摘要: Developing categorization schemes involves discovering structures in the world that support a learner's goals. Existing models of category learning, such as exemplar and prototype models, neglect role goals shaping conceptual organization. Here, clustering approach is discussed reflects joint influences environment directing acquisition. Clusters are flexible representational medium exhibits properties exemplar, prototype, rule-based models. reflect natural bundles correlated features present our environment. The model Supervised Unsupervised Stratified Incremental Adaptive Network (SUSTAIN) operates by assuming has simple structure adding complexity (i.e., clusters) when existing clusters fail to satisfy thus elicit surprise. Although simple, this operation sufficient address findings from numerous laboratory cross-cultural studies.