a bit more on our research - concept formation / learning
Several models of human learning (e.g. concept formation) have been introduced. But we consider the majority of them to have more normative (how we should think) than descriptive (how we do really think) orientation / justification. For example, although, human cognitive competence is characterized by adaptability, which manifests itself prominently in context dependent thinking skills, many computational models of human learning neglect this important characteristics, assuming that human learning involves complicated computational operations with a rigid objective (i.e., minimization of classification error). But, we believe ordinary humans are adaptive thinkers, being influenced by intra, inter, and/or extra-personal contextual factors, who would attempt to optimize (expected) situation-sensitive knowledge utility during learning. In other words, concept formation is not driven by error-minimization (i.e., accuracy maximization), but by optimization of contextually defined utility of knowledge being acquired (e.g., accuracy, simplicity, expertise, commonality, & etc...). How do we incorporate contextual factors? - see scodel or CLEAR approach. How do we make models to "realize" their intra, inter, and extra contextual factors? We do not know.
grad. students wanted
We are continuously looking for graduate students who are interested in conducting research with us on Cognitive Science and related areas (both theoretical & applied). Particularity, those who are interested in and/or familiar with mathematical modeling, computer simulation, and/or behavioral experiments are welcomed at our lab. The ideal candidates should have (a) programming experience (C++ or interpreted language - Matlab), (b) the mathematical basis to perform computational modeling, (c)the skills & knowledge to conduct empirical studies, and most importantly (d) the analytic thinking skill. If you are interesed, please contact us !