Research On Data Mining And Expert System Of Liver Disease

With the continuous development and application of computer technology and database management system, people gradually surrounded by a large amount of data, how to apply the data mining theory to the medical field become the hot spot of research. Through the analysis of medical data, mining out the hidden, valuable rules, information or knowledge is useful for the research work of the doctor's diagnosis, treatment of disease, promotion of health and so on, it has the extremely important reference value.The main research work of this paper includes three aspects as follow:(1)Studying the status of medical data mining at home and abroad, and the related issue of data mining technology, such as the classification, the clustering analysis and so on, and focusing on the existing problems of the decision tree algorithm;(2)Aimed at the problem on affecting the accuracy of decision tree algorithm, when selecting the training tuples by using random sampling method to obtain general information only but miss some valuable patterns, thus affect the accuracy of decision tree, The paper put forward the improved decision tree algorithm based on cluster sampling (IDTACS), by combining clustering algorithm to group instances and from each group according to the proportion, sampling randomly to combined into training set. The algorithm ensure that all groups have a certain number of cases in order to obtain the information of data set for maximum, improve the accuracy of decision tree, the experiment analysis results show that the improved decision tree algorithm IDTACS has good advantage in terms of accuracy.(3)After collecting the data of chronic hepatitis B patient accumulated in the hospital information management system base on electronic medical record, describing the concrete process of using the improved decision tree algorithm IDTACS to mine the hidden association rules between liver disease index and antiviral therapy in detail, then the paper designs and preliminary implies the liver diseases expert prototype system.The prototype system can preliminary establish the knowledge base according to the rules obtained by data mining, assists clinicians to make the treatment plan of patients according to their situation, it can be used to guide the treatment of liver disease, clinical teaching, scientific research work and so on.