HBTNN2e INFO
> SUBMISSION GUIDELINES
> > Topic / Content
Format
Delivery
TIMETABLE
TOPIC / CONTENT GUIDELINES FOR YOUR CONTRIBUTION
Michael A. Arbib Editor arbib@pollux.usc.edu
Prudence H. Arbib Editorial Assistant arbib@cs.ucsd.edu
The Handbook cannot be exhaustive, since the flow of research in the field is, and will continue to be, immense. But we can aim to give the reader a firm sense of the fundamental concepts and results of our field, and to provide a set of articles which give a reliable perspective on what has been achieved to date and where the field is going. The reader will then be able to approach the current literature with the tools and vocabulary we have provided. In almost all cases, the invitation is to write a review article. In general:
· the author's own work should occupy at most 30% of the text and of the list of references.
· Readers will come from a variety of disciplines, such as AI, applied mathematics, biology, computer science, electrical engineering, linguistics, neuroscience, philosophy, physics, and psychology. Please try to make all of your article accessible to this wide audience, assuming the level of a first-year graduate student.
· If this is not possible, please make at least the "Introduction" and closing "Discussion" readable without special background. Then in following sections precede specialized material (e.g., mathematical equations, neurochemistry) by brief informal motivations.
· Provisional drafts of your article can be delivered via email to Margy Avery (mavery@mit.edu) for posting on the ftp site (ftp://ftp.mitpress.mit.edu/pub/incoming/books/private/arbib).
Please ensure your drafts conform to the technical specifications for formatting.
You can download a copy of the form for request for Permission to reprint in PDF format. Make as many copies as you need, and write to the author and publisher of any books or articles from which you wish to reprint material, whether an extended quotation or a figure. Check permissions letters carefully when they are returned to make sure that stipulations regarding acknowledgments are incorporated into your manuscript. Any fees are the author's, not the editor's or publisher's, responsibility. A complete file of permissions letters should be included with the copies of your final manuscript.
SPECIFIC GUIDELINES FOR ENTRIES
Each article must balance the following requirements:
(i) It is authoritative within its own subfield yet accessible to students and experts in a wide range of other fields.
(ii) It is comprehensive, yet short enough that its concepts can be acquired in a single sitting. (Each article will be assigned a stipulated word length, which is to be strictly adhered to.)
(iii) The number of references in each article is to be kept to about 15 to give the reader a well defined and selective list of places to go to initiate further study of the given topic.
(iv) Each article is to be as self contained as possible, while providing cross references to allow readers to explore particular issues of related interest.
In preparing your article, please consider the title Handbook of Brain Theory and Neural Networkscarefully.
The term "Brain Theory" refers to any attempt to build a theory of the function of the brain or some part thereof. Such a theory may be mathematical, computational, or in a few cases purely conceptual. Articles on brain theory should be grounded in experimental data, but they should not be reviews of such data per se. Data need only be introduced to the extent that they further the presentation or critique of models.
The term "Neural Networks" refers both to biological neural networks (here the topic overlaps "Brain Theory") and to artificial neural networks. Articles may stress general perspectives or specific methods.
The term Handbook is more equivocal. The book might have been termed an Encyclopedia. However, the choice of Handbook comes with the request that authors not simply list the features of models, but also give the reader a sense of how and why they are used, and of how to evaluate the models (or the simplifying assumptions made in unifying theories), indicating both their successes and failures. The idea is to make the Handbook not only a valuable reference work but also a tool for future researchers. For example, if the article treats a specific neural network system, please include the following in your treatment:
(i) A clear statement of the problem.
(ii) State what neural network technique is used. Review the technique briefly -- perhaps with a cross-reference to another article.
(iii) State clearly the success rate of the system. If possible, benchmark it against other approaches, neural or non-neural.
(iv) Provide pointers to related applications and/or research, including discussion of biological analogs where appropriate.
Some invitees intend to write with co-authors. You are certainly welcome to have as many co-authors as you feel will strengthen the article. In particular, articles in brain theory may benefit from collaboration between a modeler and an experimentalist. However, the economics of the project are such that only one "corresponding author" of an article will receive a complimentary copy of the Handbook. To avoid confusion, official correspondence about a paper will be with that author. (If you have not already done so, please send the corresponding authors' Letter of Agreement, signed by the senior author, to
Prue Arbib
Handbook of Brain Theory and Neural Networks
8734 Dunaway Drive
La Jolla, CA 92307
Phone: 858-677-0970
Fax: 213-740-5687
September, 2001.