New Methods to Construct Membership Functions and Generate Fuzzy Classification Rules Based on the Distribution of Training Instances

• Main Authors: Yao-De Fang, 方耀德
• Other Authors: Shyi-Ming Chen
• Format: Others
• Language: en_US
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/90679754080570883784

碩士 === 國立臺灣科技大學 === 電子工程系 === 91 === The fuzzy classification system is an important application of the fuzzy set theory. Fuzzy classification systems can deal with perceptual uncertainties in classification problems. In this thesis, we propose two methods to deal with fuzzy classification problems for fuzzy classification systems. The first method deals with the Iris data classification problems based on the distribution of training instances. It can get a higher average classification accuracy rate than the existing methods. The second method constructs and tunes membership functions and generates fuzzy classification rules from training instances for handling the Iris data classification problem. It tunes the membership functions to improve the average classification accuracy rate. It can get a higher average classification accuracy rate, generate fewer fuzzy rules, and generate fewer inputs fuzzy sets in the generated fuzzy rules than the existing methods.