Study on Constructing Support Vector Machine with Granular Computing

摘要

Machine learning based on data has been a focus in the field of artificial intelligence research and application. As the process of constructing support vector machine can be abstracted as multiple granules. With the granular, the set of granules and the relationship of those granules, it can also be granulated and divided as several parts, and then be combined together as an integrity by organization. Meanwhile, based on the multiple granules and the relationship of those granules, obtained from the process of building SVM, can accelerate the process as much as possible. Finally, we focus on the training time and generalization of the large scale SVM data that are based on the granular computing.