個人化智慧型過敏資訊服務系統之設計與實現

The object of this project is to develop a personal intelligent allergic information system (PIAIS) with association inference and clustering analysis functions for allergic factors and allergic patient, respectively. PIAIS can be used to the self-allergic factors query for the public and clinical diagnostic references for the doctors or experts, respectively. In PIAIS, the Apriori algorithm is adopted to mine the association rules of the allergic factors first. Then, the allergic patients are clustered using K-means algorithm. Due to the number of the allergic factors is numerous; a feature selection method is proposed based on genetic algorithm and neural network approaches to select the most important allergic factors. Simulation result shows that the proposed PIAIS can clarify the association relationship of allergic factors and identify the cluster characteristic of allergic patients. Moreover, the cluster performance can be improved though the number of considered allergic factors is reduced largely.