A Model for Missing Value Processing in Database with Self-Organizing Map

• Main Authors: Yu-Hsiang Chiu, 邱鈺翔
• Other Authors: Chen-Chau Yang
• Format: Others
• Language: zh-TW
• Published: 2002
• Online Access: http://ndltd.ncl.edu.tw/handle/66744883922564150909

碩士 === 國立臺灣科技大學 === 電子工程系 === 90 === In this thesis, we propose a method that uses Self-Organizing Feature Map and Fuzzy Set Theory to process the data with missing value problem in database. In real world, the data we collected are not complete, and have the problem of missing value. If we analyze it directly, we usually can not get useful information, and the result of analysis will not be correct. The method we proposed is based on Self-Organizing Feature Map to analyze the data in database, and we use Fuzzy Set Theory to fuzzify the output neurons. Then we can compute values for the data with missing value. By this method, we hope to fulfill the needs of data preprocessing in Knowledge Discovery in Database, and improve the correctness of the follow-up analysis. We also implement this method and use the air quality monitoring data to justify the feasibility of the method we proposed.