Font Size: a A A

Research And Application On The Case-based Reasoning System For Judgment Of Diabetes

Posted on:2014-06-29Degree:MasterType:Thesis
Country:ChinaCandidate:Y LuoFull Text:PDF
GTID:2298330422490411Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Case-Based Reasoning Technology(CBR, Case-based reasoning) is a newproblem-solving approach is the field of artificial intelligence. The basic idea of thisapproach is to solve the current problem by using the historical problem-solvingexperience which has the highest similarity. This approach is a reasoning methodbased on historical experience. The paper solve the problems such as the limitationof experts’ working time, the experience and knowledge of the experts are not wellkept or share. The main content of the paper are as follows:First of all, the dissertationlearned the basic principle of case-based reasoning,including the case-based reasoning function model, division of research task,case-based reasoning system applied research and other developments. Then thedissertationgave the elaboration of case-based reasoning divided part such as caserepresentation, case retrieval, case reuse, case revise and case retain.Secondly, after analyzing a large number of historical diabetes cases andconsidering the characteristics of diabetes, the dissertationidentified the caserepresentation method and creates the case base.After studying the theoretical basisof case-based reasoning, the dissertationestablished a diabetes expert systemdetermines process, and introduced the details. The dissertationstudied two differentways to identify the weights: setting weights by experts and setting weights basedon the standard deviation. And the dissertationalso studied a method which did notneed to set weights, which is principal component analysis. After that, thedissertationstudied three different ways to calculate the similarity. They areEuclidean distance calculation method, improved cosine function calculationmethod and based on normalized similarity calculation method. Thedissertationtested and verified different setting weights methods.Then the dissertationgave the summary and analysis of the case-basedreasoning system for diabetes expert judgment theoretician’s framework andimplementation process, and gave the further research work in the right direction.Finally, the system was established by B/S structure, using C#programinglanguage to develop in visual studio2010.
Keywords/Search Tags:case-based reasoning, expert system, diabetes, standard deviation, similarity calculation
PDF Full Text Request
Related items