摘要: Using a recently introduced proximal support vector machine classifier [4], very fast and simple incremental (SVM) is proposed which capable of modifying an existing linear by both retiring old data adding new data. A important feature the single-pass algorithm , allows it to handle massive datasets, that huge blocks data, say order millions points, can be stored in size (n + 1), where n usually small (typically less than 100) dimensional input space resides. To demonstrate effectiveness we classify dataset 1 billion points 10-dimensional into two classes 2.5 hours on 400 MHz Pentium II processor.
参考文章(9)
Filip M. Mulier, Vladimir Cherkassky, Learning from Data: Concepts, Theory, and Methods John Wiley & Sons, Inc.. ,(1998)
Olvi Mangasarian, Generalized Support Vector Machines ,(1998)
Olvi L. Mangasarian, Yuh-Jye Lee, RSVM: Reduced Support Vector Machines siam international conference on data mining. pp. 1- 17 ,(2001)
Glenn Fung, Olvi L. Mangasarian, Proximal support vector machine classifiers knowledge discovery and data mining. pp. 77- 86 ,(2001) , 10.1145/502512.502527
O.L. Mangasarian, D.R. Musicant, Successive overrelaxation for support vector machines IEEE Transactions on Neural Networks. ,vol. 10, pp. 1032- 1037 ,(1999) , 10.1109/72.788643
Theodoros Evgeniou, Massimiliano Pontil, Tomaso Poggio, Regularization Networks and Support Vector Machines Advances in Computational Mathematics. ,vol. 13, pp. 1- 50 ,(2000) , 10.1023/A:1018946025316
Olvi L Mangasarian, David R Musicant, Lagrangian support vector machines Journal of Machine Learning Research. ,vol. 1, pp. 161- 177 ,(2001) , 10.1162/15324430152748218
Vladimir N. Vapnik, The Nature of Statistical Learning Theory ,(1995)
David R. Musicant, Olvi L. Mangasarian, Active Support Vector Machine Classification neural information processing systems. pp. 577- 583 ,(2000)