Description:Computer-based forecasting methods are no longer restricted to the traditional statistical methods. Knowledge-based approaches, which utilise the knowledge gained from human experts in their construction, and machine learning methods in which the computer learns from available examples without significant human intervention are now in general use as forecasting tools. This book examines the application of these methods to the domain of horseracing in Great Britain and overseas and provides a detailed step-by-step guide to implementing these methods. Included are instructions on the use and development of rule-based systems for horseracing, an introduction to developing knowledge-based systems including methods for handling inexact and uncertain data, a description of neural neyworks and a guide to how these can be applied to the horseracing problem and several examples illustrating the proposed methods complete with evaluation.