An Efficient Algorithm for Incremental Data Mining Based on Clustering Technique

• Main Authors: Zhi-Hang Mei, 梅志航
• Other Authors: K.M Yu
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/91586279800125311468

碩士 === 中華大學 === 資訊管理學系 === 92 === Nowadays the different techniques of data mining played an important role by widely used in many fields, such as Customer Relationship Management (CRM), Knowledge Management (KM) and behavior analysis. Ever since the implementation of the National Health Insurance, people have got a better medical environment. The medical treatments and related information have been pushed to be digitalized in the hospital record-keeping system. There are some useful medical patterns in the enormous digitalized medical record. Those patterns could help doctor with diagnosis and help the patient to better understand their disease. However, the massive medical record is accumulated too fast to be handled by human. In order to discover the hidden patterns from all the records, the data mining technique is needed. Since the electronic medical record is accumulating day by day therefore the most suitable data mining technique would be incremental data mining which can discover connected patterns. From the distinction of the incremental data mining we conceived an efficient algorithm- ICTC (Incremental Clustering Table Counting). In this thesis, we also designed some experiments for ICTC algorithm, and proved that ICTC algorithm is more efficient and stable than other incremental data mining algorithm.