New Methods for Handling Classification Problems Based on Fuzzy Entropy Measures and Fuzzy Information Gain Measures

• Main Authors: Jen-Da Shie, 謝政達
• Other Authors: Shyi-Ming Chen
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/93504009282990968312

碩士 === 國立臺灣科技大學 === 資訊工程系 === 94 === Classification techniques have been widely applied in many domains. In this thesis, we propose two new methods for handling classification problems. The first method selects feature subsets for handling classification problems based on fuzzy entropy measures focusing on boundary samples. It can deal with both numeric and nominal features. It can select relevant features to get higher average classification accuracy rates than the ones selected by the existed methods. The second method handles classification problems based on fuzzy information gain measures. It can deal with both numeric and nominal features. It can get higher average classification accuracy rates than the existed methods.