This paper has two objectives. First, we will argue that the mutability of conceptual fea- tures can be represented as a single, multiple-valued dimension. We will show that the fea- tures of a concept can be reliably ordered with respect to the degree to which people are willing to transform the feature while retaining the integrity of a representation; i.e., that a number of conceptual tasks, all of which require people to transform conceptual features, produce similar orderings. Following Medin and Shoben (1988), these tasks have in common that they ask people to consider an object that is missing a feature but is otherwise intact (e.g., a real chair without a seat)

Abstract not found

is a network model of human category learning. SUSTAIN initially assumes a simple category structure. If simple solutions prove inadequate and SUSTAIN is confronted with a surprising event (e.g. it is told that a bat is a mammal instead of a bird), SUSTAIN recruits an additional cluster to represent the surprising event. Newly recruited clusters are available to explain future events and can themselves evolve into prototypes/attractors/rules. SUSTAIN has expanded the scope of findings that models of human category learning can address. This paper extends SUSTAIN so that it can be used to account for both supervised and unsupervised learning data through a common mechanism. A modified recruitment rule is introduced that creates new conceptual clusters in response to surprising events during learning. The new formulation of the model is called uSUSTAIN for ‘unified SUSTAIN. ’ The implications of using a unified recruitment method for both supervised and unsupervised learning are discussed. Keywords:

Abstract not found

A significant component of human observational learning is the ability to focus attention toward important or relevant input features. Amechanism with this capability can serve as an inductive bias to facilitate learning in both humans and machines. Past attempts to model attentional focus for human learning have postulated a single salience value for each feature, such that features with greater salience command more attention. These models, however, assume that the feature's salience is not dependent on context, whereas studies of human attention show sensitivity to context. This paper presents a mechanism for contextually focused attention in observational learning. 1

Investigations of the degree to which people neglect or use base rates typically require subjects to make a judgment based on presumptive integrations of base rates and likelihood ratios. The present paper deals with a logically prior issue, whether people understand what data are needed to constitute a proper base rate. The method, which we will call the Partial Information Paradigm, has subjects select data relevant to, for example, diagnosis of a disease, D, based on a symptom, S. The question is whether subjects select those frequencies of cases for which information about the presence or absence of D is available, but for which information about the presence or absence of S is not. Only the former frequencies are relevant to the estimation of the base rate of D, hence to the probability of D given S. Six experiments are reported. Four experiments ask subjects to select those frequencies relevant to diagnosis, one of which also had subjects select frequencies relevant to prediction...

The aim of this paper is to present an evolutionary framework for categorization. Evolution needs an evaluation mechanism to work, and it is argued that primary values that the organism needs for its survival -- such as food, mates for reproduction, and shelter -- can drive the evolution of categories. Sensory stimulation is needed to build up the cognitive apparatus, but cannot in itself provide the evaluation mechanism for evolution. Categorization constrained by values will be dependent on the availability of sensory information, and its power as predictive of values. As perception and categorization are tied to the actions of the organism, it is argued that the unit of perception should be seen as larger than the usual singledimension stimulus, and evidence is reviewed to support this claim. Covarying stimuli will also provide a much greater predictive power than single-dimension stimuli alone.

Abstract not found

The present research explored the role of working memory (WM) in unsupervised category learning, learning without an external tutor or even knowing that categories exist, by investigating its role using a pattern-sequence manipulation. A pattern-sequence manipulation compares learning when items from categories are presented together (blocked) versus when the items are presented in random order (mixed). Experiment 1 extended the pattern-sequence manipulation to assess category knowledge separate from paired-associate learning. Participants performed equally well on new and studied items, supporting the hypothesis that the pattern-sequence manipulation results in the acquisition of category information, not simply memory for item-feature associations. Experiment 2 introduced a WM factor, administering the method used in Experiment 1 to a group of high and low WM span participants. High WM span was predicted to interact with the pattern-sequence effect to produce greater learning when the items were blocked than mixed. There was reliable support for a role of WM span in the discovery and acquisition of category knowledge, but this role was different from the one hypothesized. The high WM span participants exhibited higher overall accuracies than the low WM span participants. This result supports a role for WM in unsupervised category learning, but did not benefit more from the pattern-sequence effect than did the low WM span participants as predicted. Implications for theories of category learning and WM are discussed.

Abstract not found