Neural-Network Models of Cognition
John W. Donahoe, Vivian Packard Dorsel (eds), 1997

Chapter 1. The Necessity of Neural Networks
John W. Donahoe

The goal of this volume is to provide a foundation for a natural science-based understanding of human behavior, including the complex phenomena of "mind" -- perception, thought, memory, and language.

Darwin's work teaches us that the function of structure determined the fate of structure and that complex structure/function arose as the emergent product of lower-level processes acting over time. No appeal to higher-level principles was required. Since selection only acted in the past, selection has "prepared" us to live in the past, not the future. The conventional attributes of mind -- foresight, intelligence, reason, and the like -- are comforting illusions which endure only insofar as future conditions are similar to past conditions.

Acceptance of a natural science account of human morphology has come slowly to scientists and even more slowly to laypersons. Acceptance of this account of human behavior has been even more slowly still. Part of the reluctance of scientists to accept natural evolution can be understood in that a science of genetics developed only as a result of Mendel's work in the early 1900s. Secondly, Fisher, Haldane and Wright established the formal techniques of population genetics in the 1920s. These three components -- a functional principle, a biological mechanism, and a formal technique -- were necessary preludes to the general acceptance of evolution though natural selection.

A parallel is presented here between the natural selection effects which allow structure to evolve into complex patterns and behavior selection effects, acting within the lifetime of an individual, which allow the evolution of complex behavioral patterns in the individual. Because such effects act more quickly than the natural selection of structure, they can be observed in the laboratory. Ivan Pavlov and Edward Thorndike independently began a search for the principles of such behavioral selection.

Two fundamental factors have been identified that, together, allow experience to change behavior. First, the critical events must occur very close in time to be effective -- only a few seconds apart at most. The second factor, discovered more recently, is that the behavioral selection would occur only if that selection would result in a change in behavior. For example, if salivation was already evoked by a prior light-food pairing, then addition of a tone simultaneously with the light would not cause the adaptation of salivation on presentaion of the tone.

Many of the papers in this volume trace the development of neural network architectures, showing how such networks can be devised so as to exhibit the necessary patterns of activity needed to produce the desired bahavior and also to conform to formal principles needed to characterize their application as an explanation for the complex behavior.

Part One. Network Architecture and Neuroanatomy

Chapter 2. Progenitor Cells of the Mammalian Forebrain:Their Types and Distribution
Marla B. Luskin

Chapter 3. A Statistical Framework for Presenting Developmental Neuroanatomy
Stephen L. Senft

Chapter 4. Evolving Artificial Neural Networks in Pavlovian Environments
José E. Burgos

Part Two. Neural Plasticity

Chapter 5. Principles of Neurotransmission and Implications for Network Modeling
Jerrold S. Meyer

Chapter 6. Cellular Mechanisms of Long-Term Potentiation: Late Maintenance
Uwe Frey

Chapter 7. Temporal Information Processing: A Computational Role for Paired-Pulse Facilitation and Slow Inhibition
Dean V. Buonomano and Michael M. Merzenich

Part Three. Perceiving

Chapter 8. Development and Plasticity of Neocortical Processing Architectures
Wolf Singer

Chapter 9. Inferotemporal Cortex and Object Recognition
Keiji Tanaka

Chapter 10. Sparse Coding of Faces in a Neuronal Model: Interpreting Cell Population Response in Object Recognition
Arnold Trehub

Chapter 11. Structure and Binding in Object Perception
John E. Hummel

Chapter 12. A Neural-Network Approach to Adaptive Similarity and Stimulus Representations in Cortico-Hippocampal Function
Mark A. Gluck and Catherine E. Myers

Part Four. Behaving

Chapter 13. Motor Cortex: Neral and Computational Studies
Apostolos P. Georgopoulos

Chapter 14. Selectionist Constraints on Neural Networks
David C. Palmer

Chapter 15. Analysis of Reaching for Stationary and Moving Objects in the Human Infant
Neil E. Berthier

Chapter 16. Reinforcement Learning of Complex Behavior through Shaping
Vijaykumar Gullapalli

Part Five: Reinforcement Learning

Chapter 17. Adaptive Dopaminergic Neurons Report the Appetitive Value of Environmental Stimuli
Wolfram Schultz

Chapter 18. Selection Networks: Simulation of Plasticity through Reinforcement Learning
John W. Donahoe

Chapter 19. Reinforcement Learning in Artificial Intelligence
Andrew G. Barto and Richard S. Sutton

Chapter 20. The T-D Model of Classical COnditioning: Response Topography and Brain Implementation
J. W. Moore and J-S Choi

Chapter 21. Biological Substrates of Predictive Mechanisms in Learning and Action Choice
P. Read Montague

Chapter 22. The Role of Training in Reinforcement Learning
Jeffrey A. Clouse

Part Six. Complex Behavior -- Language

Chapter 23. Functional Brain Imaging and Verbal Behavior
Marcus E. Raichle

Chapter 24. Neural Modeling of Learning in Verbal Response-Selection Tasks
Vijaykumar Gullapalli and Jack J. Gelfand

Chapter 25. Serial Order: A Parallel Distributed Processing Approach
Michael I. Jordan

Chapter 26. Connectionist Models of Arbitrarily Applicable Relational Responding: A Possible Role for the Hippocampal System
Dermot Barnes and Peter J. Hampson

Chapter 27. A Recurrent-Network Account of Reading, Spelling, and Dyslexia
Guy C. Van Orden, Anna M. T. Bosman, Stephen D. Goldinger, and William T. Farrar, IV

Top of Page | NNMC Opinion | Sort by Topic | Sort by Title | Sort by Author