The Research Of Diagnosis On A H1N1Influenza On Case-Based Reasoning

The thesis comes from influenza H1N1diagnosis of some typical cases of influenza diagnosis and diagnosis-based artificial intelligence expert system in the actual application of the medical field. Papers sub-theory and application, combined with CBR theory to the diagnosis of influenza A H1N1influenza analysis and research, this paper presents a auxiliary system solution.CBR paper described the basic concepts and principles, combined with the diagnosis of influenza A H1N1influenza knowledge, describes the core of CBR system which links with a number of methods:case representation and storage, case retrieval and matching, case changes and learning case library maintenance, add and delete instances.The clustering technique applied to the CBR system, using improved CURE_KNN algorithm, examples of case library in the cluster. In the case of that used in object-oriented approach, effective solution to the issue of knowledge representation; by way of building the standard hierarchical case library; in case retrieval, and the combination of nearest neighbor method and the reduce in different levels according to some tactic is employed in case retrieval, to develop the retrieval efficiency and reduce the process of retrieval we use the conception of attributes space in the retrieval process; in case library maintenance, increase based on clustering of case,case deletion strategy, significantly improved the speed and quality of case retrieval. Verified by the system running the test the effectiveness of these methods.