Forecasting Methods For Horseracing

systems. Hopefully the reader will be able to recreate the methods presented, and more importantly, generate new, unique, forecasting systems that produce even more impressive results by constructing more detailed approaches using the theories expounded in this text. To assist with the development of neural network approaches, a back-propagation algorithm is given in the appendix which can be easily implemented in any programming language. Alternatively, a basic neural network simulation program, details of which are given in appendix A4, which can be used to train small neural network applications....

_______________________ References

1 Searching For Certainty- what science can know about the future, by J. L Casti, Scribners, London, England, 1992.

2 An Application of a Multiple Neural Network Learning System to Emulation of Mortgage Underwriting Judgements, by E. Collins, S. Ghosh and C L. Scofield, in IEEE International Conference on Neural Networks, San Diego, CA, II, pp. 459-466, 1988.

3 Why Artificial Intelligence needs an EmpiricalFoundation, byN. E.Sharkeyand G.D.A.Brown,in Artificial Intelligence: Principles and Applications, M.Yazdani (ed.), pp. 267-293,1986.

4 Scientific Handicapping: Tested Ways to Win at the Track, by I. S.Cohen and G. D. Stephens, Prentice-Hall, Englewood Cliffs, NJ. 1963.

5 Connectionism and the Mind: Introduction to Parallel Processing in Networks.by A. Bechtel and A. Abrahamsen, Oxford, England: Blackwell, 1991.

6 Hybrid Architectures for Intelligent Systems, by A. Kandel, CRCP, London, May 1992.

19