Title :
Fast linear stationary methods for automatically biased support vector machines
Author :
Lai, D. ; Palaniswami, M. ; Mani, N.
Author_Institution :
Dept. of Electr. & Comput. Syst. Eng., Monash Univ., Clayton, Vic., Australia
Abstract :
We present a new training algorithm, which is capable of providing fast training for a new automatically biased SVM. We compare our algorithm to the well-known sequential minimal optimization (SMO) algorithm. We then show that this method allows for the application of acceleration methods which further increases the rates of convergence.
Keywords :
convergence of numerical methods; extrapolation; iterative methods; learning (artificial intelligence); optimisation; support vector machines; acceleration methods; automatically biased SVM; convergence; extrapolation; iterative methods; linear stationary methods; sequential minimal optimization algorithm; support vector machines; training algorithm; Acceleration; Australia; Jacobian matrices; Machine learning; Neural networks; Optimization methods; Pattern recognition; Risk management; Support vector machines; Systems engineering and theory;
Conference_Titel :
Neural Networks, 2003. Proceedings of the International Joint Conference on
Print_ISBN :
0-7803-7898-9
DOI :
10.1109/IJCNN.2003.1223725
