TITLE:
LP-SVR Model Selection Using an Inexact Globalized Quasi-Newton Strategy
AUTHORS:
Pablo Rivas-Perea, Juan Cota-Ruiz, J. A. Perez Venzor, David Garcia Chaparro, Jose-Gerardo Rosiles
KEYWORDS:
Hyper-Parameter Estimation; Support Vector Regression; Machine Learning; Data Mining
JOURNAL NAME:
Journal of Intelligent Learning Systems and Applications,
Vol.5 No.1,
February
22,
2013
ABSTRACT:
In this paper we study the problem of model selection for a linear programming-based support vector machine for regression. We propose generalized method that is based on a quasi-Newton method that uses a globalization strategy and an inexact computation of first order information. We explore the case of two-class, multi-class, and regression problems. Simulation results among standard datasets suggest that the algorithm achieves insignificant variability when measuring residual statistical properties.