| Summary: | 碩士 === 國立中興大學 === 應用數學系所 === 100 === The analyses of ordinal data have received increasing attention in many research areas such as medicine and image retrieval. In this study, we use a paired-sample approach to classify the ordinal data by the support vector machine (SVM) in combination with the cumulative logit model (CLM). We explain how to determine the parameters of ordinal regression model by this method. The sequential minimal optimization (SMO), which is most commonly used for optimizing the SVM, and an improved version of the algorithm used in this study are discussed. Finally, we apply the method to artificial data, real medicine data and image data to evaluate its classification performances.
|
|---|
