Kernel logistic regression (KLR) is a powerful and flexible classification algorithm, which possesses an ability to provide the confidence of class prediction., However, its training - typically carried out by Newton's method or quasi-Newton methods - is rather time-consuming. In this paper, we propose an alternative probabilistic classification algorithm called Least-Squares Probabilistic Classifier (LSPC). The solution of LSPC can be computed analytically just by solving a system of linear equations, so LSPC is computationally efficient and stable. Through experiments, we show that the computation time of LSIPC is faster than that of KLR by the factor 100 with comparable accuracy.
展开▼
