Abstract:In this paper, a novel non-parametric method to map SVM outputs into posterior probabilities is proposed. Firstly, we use Support Vector Regression to fit the class conditional probabilistic density. Then Baye's rule is used to produce the posterior probabilities for a given sample. As for multi-class classification problem, we use Pairwise Coupling method. This approach is used for handwritten digital recognition problem and performs very well, which illuminates the approach proposed in this paper is efficient and effective.

Non-Parametric Estimation Of Svm Probabilistic Outputs And Its Application On Handwritten Digital Recognition