Design And Realization For Online Diagnoses System Based On Medical Data Mining

Though the socio-economic and material life are improved continuously, the threat of chronic non-communicable diseases to humans (especially the elderly)’s health is still growing. To analyze and extract information from the patients data is useful for doctors to carry out targeted treatment, and is also more conducive to the rehabilitation of patients. It is an important task in the field of data analysis to design and develop a data classification system, which is easy to use, with high efficiency and good effective.First, this paper introduces the feature selection and data classification algorithms, focusing on the discrete data and inconsistent data rate, feature selection methods is based on inconsistent rate, and data classification method is based on inconsistent rate;Second, this paper introduces the pre-processing of multi-dimensional medical data, including data format conversion, data cleansing, and display the results of this article medical data preprocessing; Third, this paper analysis the condition of the multi-dimensional medical data mining, introduce the set of disease classification, demonstrats the feature selection and case classification of medical data.The last part describes completion of a Weka, which realizes the functions, such as rapid classification of data and classification results visualization. In addition, design an online condition data analysis system, using Java technology and Weka newly developed algorithm, which is based on the feature selection and rule extraction. It realizes data mining, such as automatic classification of data, data mining,data forecast and meet the users’ needs of online data mining and data category forecast. Also, this paper give write the test script for the online condition data analysis system, to verify the efficiency and practicality of the system.