Optimizing working sets for training support vector regressors by Newton's method

• Published in: 2015 International Joint Conference on Neural Networks (IJCNN) pp. 1 - 8
• Main Author: Abe, Shigeo
• Format: Conference Proceeding Journal Article
• Language: English
• Published: IEEE 01.07.2015
• Subjects: Conferences; Derivatives; Integrated circuits; Mathematical analysis; Neural networks; Newton methods; Optimization; Training; Vectors (mathematics)

In this paper, we train support vector regressors (SVRs) fusing sequential minimal optimization (SMO) and Newton's method. We use the SVR formulation that includes the absolute variables. A partial derivative of the absolute variable with respect to the associated variable is indefinite when th...