The papers that appear in this volume are refereed versions of presenta tions made at the third Neural Computation and Psychology Workshop, held at Stirling University, Scotland, from 31 August to 2 September 1994. The aim of this series of conferences has been to explore the interface between Neural Computing and Psychology: this has been a fruitful area for many researchers for a number of reasons. The development ofNeural Computation has supplied tools to researchers in Cognitive Neuroscience, allowing them to look at possible mechanisms for implementing theories which would otherwise remain 'black box' techniques. These theories may be high-level theories, concerned with interaction between a number of brain areas, or low-level, describing the way in which smaller local groups of neurons behave. Neural Computation techniques have allowed computer scientists to implement systems which are based on how real brains appear to function, providing effective pattern recognition systems. We can thus mount a two-pronged attack on perception. The papers here come from both the Cognitive Psychology viewpoint and from the Computer Science viewpoint: it is a mark of the growing maturity of the interface between the two subjects that they can under stand each other's papers, and the level of discussion at the workshop itself showed how important each camp considers the other to be. The papers here are divided into four sections, reflecting the primary areas of the material.

Klappentext

The papers that appear in this volume are refereed versions of presenta­ tions made at the third Neural Computation and Psychology Workshop, held at Stirling University, Scotland, from 31 August to 2 September 1994. The aim of this series of conferences has been to explore the interface between Neural Computing and Psychology: this has been a fruitful area for many researchers for a number of reasons. The development ofNeural Computation has supplied tools to researchers in Cognitive Neuroscience, allowing them to look at possible mechanisms for implementing theories which would otherwise remain 'black box' techniques. These theories may be high-level theories, concerned with interaction between a number of brain areas, or low-level, describing the way in which smaller local groups of neurons behave. Neural Computation techniques have allowed computer scientists to implement systems which are based on how real brains appear to function, providing effective pattern recognition systems. We can thus mount a two-pronged attack on perception. The papers here come from both the Cognitive Psychology viewpoint and from the Computer Science viewpoint: it is a mark of the growing maturity of the interface between the two subjects that they can under­ stand each other's papers, and the level of discussion at the workshop itself showed how important each camp considers the other to be. The papers here are divided into four sections, reflecting the primary areas of the material.

Inhalt
Cognition.- Symbolic and Subsymbolic Approaches to Cognition.- Mapping Across Domains Without Feedback: A Neural Network Model of Transfer of Implicit Knowledge.- Modelling Reaction Times.- Chunking: An Interpretation Bottleneck.- Learning, Relearning and Recall for Two Strengths of Learning in Neural Networks 'Aged' by Simulated Dendritic Attrition.- Perception.- Learning Invariances via Spatio-Temporal Constraints.- Topographic Map Formation as Statistical Inference.- Edge Enhancement and Exploratory Projection Pursuit.- The Perceptual Magnet Effect: A Model Based on Self-Organizing Feature Maps.- How Local Cortical Processors that Maximize Coherent Variation Could Lay Foundations for Representation Proper.- Audition and Vision.- Using Complementary Streams in a Computer Model of the Abstraction of Diatonic Pitch.- Data-Driven Sound Interpretation: Its Application to Voiced Sounds.- Computer Simulation of Gestalt Auditory Grouping by Frequency Proximity.- Mechanisms of Visual Search: An Implementation of Guided Search.- Categorical Perception as an Acquired Phenomenon: What are the Implications?.- Sequence Learning.- A Computational Account of Phonologically Mediated Free Recall.- Interactions Between Knowledge Sources in a Dual-Route Connectionist Model of Spelling.- Author Index.