References — Some of the source material

by Ben Best


NEUROPHILOSOPHY by Patricia Churchland (1986) is an excellent introduction to both technical and philosophical issues concerning the relationship between mind and brain. WET MIND by Stephen Kosslyn and Oliver Koenig (1992) is a layperson's-language general overview of scientific questions concerning the way the brain produces the mind. NATURE'S MIND by Michael Gazzaniga (1992) is also intended as a layperson's introduction, but it focuses on a selection of specific issues rather than giving a general overview.

PRINCIPLES OF NEURAL SCIENCE, edited by Eric Kandel, (1991) is the best general reference I have had for all topics apart from neural networks. Nonetheless, I have invariably found that it is a "good place to start", but that it is lacking in detail. NEUROBIOLOGY (Second Edition) by Gordon Shepherd (1988), FROM NEURON TO BRAIN by John Nicholls, (1992) and NEURONS AND NETWORKS by John Dowling (1992) all supplement Kandel as sources of general information and diagrams. THE CENTRAL NERVOUS SYSTEM by Per Brodal (1992) has been a valuable source of excellent diagrams and functional anatomy of brain structures.

For cellular neurophysiology I have gotten much material from AN INTRODUCTION TO MOLECULAR NEUROBIOLOGY, edited by Zach Hall (1992) — although I suspect that the authors were intent on dumping every piece of knowledge they have, without the least interest-in or understanding-of what is relevant and what is not. THE NEURON by Irwin Levitan and Leonard Kaczmarek (1991) is better for clear and meaningful exposition of cell and molecular biology questions (although it misses much detail). The PROGRESS IN NEUROSCIENCE reprint of Scientific American articles, and the October 1985 issue of Scientific American also assisted in clarifying many cell biology issues for me. MOLECULAR CELL BIOLOGY by James Darnell, (1986) has also been a helpful reference.

For the topics of learning and memory, THE NEUROSCIENCE OF ANIMAL INTELLIGENCE by Euan Macphail (1993) stands head and shoulders above any other books I have seen — both in technical detail and clarity of exposition. THE NEUROBIOLOGY OF MEMORY by Yadin Dudai (1989), on the other hand, contains a wealth of information presented in a terribly (and unnecessarily) difficult manner. BRAIN MECHANISMS OF PERCEPTION AND MEMORY, edited by Taketoshi Ono (1993) is a very good collection of articles by top researchers in the field. THE BIOLOGY OF EMOTIONS by Jean-Didier Vincent (1986 — 1990 translation by John Hughes) is an extremely entertaining, yet technically good, exposition on a topic that is rarely described in such detail.

PARALLEL DISTRIBUTED PROCESSING, edited by James McClelland and David Rumelhart (1986) is a hefty two-volume tome that launched the current renaissance of research into mathematical, computational, psychological and biological issues concerning neural networks. But for both depth and clarity, I think the newer books are better. APPRENTICES OF WONDER by William Allman (1989) is the most determinedly nontechnical, popularized introduction to neural networks I have seen. A PRACTICAL GUIDE TO NEURAL NETS by Marilyn Nelson and W. Illingworth (1991) is also a popularized account, although it attempts to deal with computer issues while avoiding mathematics. THE COMPUTATIONAL BRAIN by Patricia Churchland and Terrance Sejnowski (1992) gives a decent overview of computational, biological and psychological issues associated with neural networks, but I had hoped for something better from those two authors. THE COMPUTING NEURON, edited by Richard Durbin, (1989) is an interesting collection of articles concerning biological neural network issues — and it is not out-of-date. For the serious mathematician I would recommend NEURAL NETWORKS: A COMPREHENSIVE FOUNDATION by Simon Haykin (1994).

In my opinion, the very best theoretical introductions to neural networks are NEURAL COMPUTING: AN INTRODUCTION by R. Beale and T. Jackson (1990), and NEURAL NETWORK ARCHITECTURES by Judith Dayhoff (1990). The best practical introduction is UNDERSTANDING NEURAL NETWORKS (Volume 1) by Maureen Caudill and Charles Butler (1992). NEURAL COMPUTING gives a good theoretical overview, while carefully sequestering the mathematical and computational sections for those who wish to skip them. NEURAL NETWORK ARCHITECTURES does include mathematics along with explanations, but these explanations tend to be simple and clear. UNDERSTANDING NEURAL NETWORKS is a workbook of exercises and a diskette, which are invaluable to anyone who becomes frustrated with all the theoretical descriptions of neural networks in the absence of practical examples. I very highly recommend these books for anyone not afraid of computers or simple mathematics, and who is wanting to learn about neural networks. Volume 2 of UNDERSTANDING NEURAL NETWORKS, while of some interest, is of marginal value compared to Volume 1. Caudill and Butler also wrote an expository text, NATURALLY INTELLIGENT SYSTEMS (1990), but I did not find this to be very lucid or informative. NEURAL NETWORK PC TOOLS, edited by Russell Eberhart and Roy Dubbins (1990) gives a good coverage of practical computer-related issued involved in implementing neural network software.