The Computational Brain, 25th Anniversary Edition

by Patricia S. Churchland, Terrence J. Sejnowski

Overview

An anniversary edition of the classic work that influenced a generation of neuroscientists and cognitive neuroscientists.

Before The Computational Brain was published in 1992, conceptual frameworks for brain function were based on the behavior of single neurons, applied globally. In The Computational Brain, Patricia Churchland and Terrence Sejnowski developed a different conceptual framework, based on large populations of neurons. They did this by showing that patterns of activities among the units in trained artificial neural network models had properties that resembled those recorded from populations of neurons recorded one at a time. It is one of the first books to bring together computational concepts and behavioral data within a neurobiological framework. Aimed at a broad audience of neuroscientists, computer scientists, cognitive scientists, and philosophers, The Computational Brain is written for both expert and novice. This anniversary edition offers a new preface by the authors that puts the book in the context of current research.

This approach influenced a generation of researchers. Even today, when neuroscientists can routinely record from hundreds of neurons using optics rather than electricity, and the 2013 White House BRAIN initiative heralded a new era in innovative neurotechnologies, the main message of The Computational Brain is still relevant.

Product Details

ISBN-13:	9780262533393
Publisher:	MIT Press
Publication date:	11/04/2016
Series:	Computational Neuroscience Series
Edition description:	25th Anniversary Edition
Pages:	568
Sales rank:	834,744
Product dimensions:	6.90(w) x 9.90(h) x 1.20(d)
Age Range:	18 Years

About the Author

Patricia S. Churchland is President's Professor of Philosophy Emerita at the University of California, San Diego, and Adjunct Professor at the Salk Institute for Biological Sciences. She is the author of many books, including Neurophilosophy and Brain-Wise (both published by the MIT Press).

Terrence J. Sejnowski holds the Francis Crick Chair at the Salk Institute for Biological Studies and is a Distinguished Professor at the University of California, San Diego. He was a member of the advisory committee for the Obama administration's BRAIN initiative and is President of the Neural Information Processing (NIPS) Foundation. He is the author of The Deep Learning Revolution (MIT Press) and other books.

Table of Contents

Series Foreword vii

25th Anniversary Preface ix

Preface xvii

1 Introduction 1

2 Neuroscience Overview 17

Introduction 17

Levels in nervous systems 18

Structure at various levels of organization 27

A short list of brain facts 48

3 Computational Overview 61

Introduction 61

Looking up the answer 69

Linear associators 77

Constraint satisfaction: Hopfield networks and Boltzmann machines 82

Learning in neural nets 96

Competitive learning 102

Curve fitting 105

Feedforward nets: Two examples 107

Recurrent nets 115

From toy world to real world 125

What good are optimization procedures to neuroscience? 130

Models: Realistic and abstract 136

Concluding remarks 137

4 Representing the World 141

Introduction 141

Constructing a visual world 142

Thumbnail sketch of the mammalian visual system 148

Representing in the brain: What can we learn from the visual system? 157

What is so special about distribution? 163

World enough and time 174

Shape from shading: A neurocomputational study 183

Stereo vision 188

Computational models of stereo vision 199

Hyperaciiity: From mystery to mechanism 221

Vector averaging 233

Concluding remarks 237

5 Plasticity: Cells, Circuits, Brains, and Behavior 239

Introduction 239

Learning and the hippocampus 243

Donald Hebb and synaptic plasticity 250

Memories are made of this: Mechanisms of neuronal plasticity 254

Cells and circuits 281

Decreasing synaptic strength 289

Back to systems and behavior 295

Being and timing 305

Development of nervous systems 307

Modules and networks 316

6 Sensorimotor Integration 331

Introduction 331

LeechNet 341

Computation and the vestibulo-ocular reflex 353

Time and time again 379

The segmental swimming oscillator 388

Modeling the neuron 399

Concluding remarks 411

7 Concluding and Beyond 413

Afterword 425

Appendix: Anatomical and Physiological Techniques 427

Permanent lesions 427

Reversible lesions and microlesions 430

Imaging techniques 432

Gross electrical and magnetic recording 437

Single-unit recording 440

Anatomical tract tracing 442

Notes 445

Glossary 457

References 479

Index 525

What People are Saying About This

From the Publisher

The authors have successfully integrated a number of diverse disciplines into a coherent picture of the field. The Computational Brain is a major contribution.

— Carver Mead , Gordon and Betty Moore Professor of Computer Science, California Institute of Technology

The Computational Brain describes the new style of brain-like computation, based on neural nets, as it applies to biological systems. It is written in a lively, readable style, assisted by many illustrations, yet it does not evade the many intellectual and computational problems involved. It is unique in viewing the subject from such a wide perspective while presenting detailed examples to illustrate the present state of the art. I strongly recommend it to all those interested in how nervous systems work, including the behavior of our own brain.

— Francis Crick , The Salk Institute

There is no equivalent to The Computational Brain —a unique synthesis of the fast expanding field of neural model. A rewarding experience to read.

— J. P. Changeux , Institut Pasteur

The Computational Brain documents a revolution now occurring in the neurosciences. For the past two centuries, the mainstream approach to brain and behavior has been biomedical, usurped from philosophers first by nineteenth-century neurologists, then by psychiatrists and behavioral psychologists, and now by neuropharmacologists and by cell and molecular biologists. Since the publication of Rumelhart and McClelland's Parallel Distributed Processing: Explorations in the Microstructure of Cognition in 1986, the current is slowly turning back to epistomologists whose central paradigm is communications and information processing. The Computational Brain presents contemporary neuroscience panoramically from this revolutionary point of view. It is essential reading for anyone interested in neuroscience, which has been called the last frontier of biology, not only because it explains the relevance of computation to neurobiology with extraordinary clarity, but also because the book shows unambiguously that both frontier and center are moving.

— James H. Schwartz, MD, PhD , Center for Neurobiology & Behavior, Columbia University

Editorial Reviews

Computational neuroscience is an approach to understanding the information content of neural signals by modeling the nervous system. In this well written, lively text Churchland and Sejnowski address the foundational ideas of this emerging field, examine a diverse range of neural network models, and consider future directions. Annotation c. Book News, Inc., Portland, OR (booknews.com)

Booknews

...[T[he volume stands as a significant achievement in bringing the technically challenging world of distributed modeling to the non-specialist reader, and in intelligently integrating PDP modeling into neurophysiology.
— Psyche

Bruce Bridgeman

This attractive and well-illustrated volume falls somewhere between a trade book and a textbook, with a style well suited for the Scientific American reader, as well as the active scientist, who may know something of either computer science or neuroscience but welcomes a crisp narrative that includes the necessary background from each discipline. . . . The reader will be well rewarded who seeks to understand, from well-chosen examples, how to merge the analysis of neuroscientific data with the developments of computational principles.

— Michael A. Arbib , Science

From the Publisher