The Model of Disease Diagosis on Cow Based on ANN and GA

The progress on artificial intelligence research has greatly improved the development of intelligent disease diagnosis technology. The birth of MYCIN achieved great success and gave rise to the ongoing emergence of the disease diagnosis system that is based on the Expert System(ES). However, the ES’s inherent drawbacks, such as the difficulty of acquiring knowledge and the complexity of disease diagnosis, made the disease diagnosis that relies on exclusive expert system very limited. The Artificial Neural Network’s simultaneousness, tolerance, and fuzzy inference resolve this contradiction in a timely manner. Integrating with the Expert System to build the Intelligent Disease Diagnosis system has become the main trend. This article briefly introduces the Expert system and ANN technology, studies the feasibility and mode of the integration of these twos, explores the type of Neural Network that is suitable for disease diagnose, and sets up a cow disease diagnose RBF network model, which is optimized by Genetic Algorithm. The experiment has proven that this model enjoys high diagnosis accuracy and an excellent capacity of generalization. This article was sponsored by Project 863 “Intelligent Agriculture Information Processing System – Chongqing Demonstration Zone” (Ref. No. 2001AA115320) and Chongqing University Teachers’ sponsored project “Cow Disease Diagnosis Expert System Based on Neural Network and Genetic Algorithm”.