Abstract: Sometimes we make accurate numerical judgments based on probability estimation; at other times our judgments are biased by the way the information is framed. Sometimes our preferences are internally consistent at a logical level; at other times our emotional biases generate internal contradictions. Therefore, an accurate neural network model of human decision making needs to be able to encompass both rational and irrational processes within the same system. The brain's executive system is complex enough to encompass all these processes, and definable roles are emerging for particular subregions of the prefrontal cortex. Understanding decision making, and the interactions of decision making with memory and attention, requires getting beyond a simple dichotomy of "rational versus emotional" or "analytical versus intuitive." How we make judgments and choices depends heavily on how we select what is relevant out of the information with which we are presented. This is not an issue of limited memory capacity but an issue of how we extract and categorize the "gist" of a situation. Contextual shifts can alter gist selection in ways that may or may not be optimal or adaptive. Much of the versatility of human decision making is captured by a modified adaptive resonance network with an attribute-selective matching algorithm and affective comparators.
Biography: Daniel S. Levine is Professor of Psychology at the University of Texas at Arlington. He is a Fellow of the International Neural Network Society (INNS), having served on the Board of Governors of INNS from 1995 through 2008, and as President of INNS in 1998. He was Program Chair of the 2005 International Joint Conference on Neural Networks in Montreal, and the 2008 International Conference on Intelligent Computing in Shanghai. He was a plenary speaker at the IEEE Conference on Knowledge Interactive Multi-Agent Systems in Waltham, MA, USA, in 2007, and the International Workshop on Systems, Signals and Image Processing in Chalkida, Greece, in 2005. He is the author of a graduate textbook, Introduction to Neural and Cognitive Modeling (2nd edition, Lawrence Erlbaum Associates, 2000). His recent work has focused on neural network modeling of higher cognitive function, cognitive-emotional interactions, human decision making, and brain executive functions. In addition to computational modeling he supervises a psychology laboratory that performs experiments on emotional influences in decision making. He received his PhD in Applied Mathematics at MIT under Professor Stephen Grossberg, and had a previous appointment as Associate Professor of Mathematics.