By submitting this extended abstract to 8NSCO, I certify that (1) The content and originality of this article is the responsibility of me and the other authors of the article. (2) The other authors of the article have agreed to send this article to 8NSCO. Multi-class Support Vector Machine via Probabilistic Approach Tara Mohammadi1∗, Sohrab Effati2 1Department of Statistics, Faculty of Mathematical Sciences, Ferdowsi University of Mashhad, Mashhad, Iran, 2Department of Applied Mathematics, Faculty of Mathematical Sciences, Ferdowsi University of Mashhad, Mashhad, Iran, tara.mohammadi@alumni.um.ac.ir, s-effati@um.ac.ir Abstract. Although there are methods for the multi-class support vector machine (SVM). The existing models are still based on deterministic circum- stances. In real life, we deal with data that is not entirely accurate due to inherent errors. However, in this paper, we present a method for random data with probabilistic constraints that improves the accuracy of the model compared to models based on fixed and exact data. The presented model is for a random input vector with probabilistic constraints. Using the statistical methods and mathematical expectations and their estimation, we removed the constraints from the probabilistic states. In the simulation section, we generated arbitrary samples using the Monte Carlo method and trained the data. Experiments have been conducted on one dataset available in the data repository, and we show that the proposed model is more efficient and accurate than the model based on detailed data.

Keywords