Comparative Analysis Of Optimization Techniques On Multi-class SVM

Support Vector Machine is the most efficient algorithm in machine learning for classification problems. The data generated these days requires Multi-classification and SVM’s are used for doing the same. To improve the performance and reduce the time complexity of the multi-class SVM, different optimization techniques have been used. This project aims to compare these optimization techniques on SVM on different parameters like accuracy and time. The comparative analysis has been carried out on optimization techniques which are the current state-of-art; Sequential Minimum Optimization and Dual Decomposition.