Support Vector Machine (SVM) is a machine learning classification technique that supports binary classification. In the recent years, efforts are made to extend the SVM algorithm to support multiple classifications. This paper presents a SVM based multi-knowledge-based system (SMK) design that supports multiple classifications. The proposed design is successfully tested on a classification problem. The benchmark car evaluation dataset from UCI machine learning repository is used for training and testing the SMK. The SMK shows good performance on this classification and shows good promise for the future.
|Number of pages
|Procedia Computer Science
|Published - 2017
|7th International Conference on Advances in Computing and Communications, ICACC 2017 - Kochi, Kerala, India
Duration: 22 Aug 2017 → 24 Aug 2017