| 摘要 |
One embodiment of the present invention provides a system that performs parallel grouping decomposition to facilitate expedited training of a support vector machine (SVM). During operation, the system receives a training dataset comprised of data vectors. The system then determines whether any data vector in the dataset violates conditions associated with a current SVM. Next, the system divides the violating data vectors into a number of subsets, thereby allowing parallel SVM training for each subset. The system subsequently builds an independent SVM for each subset in parallel based on the current SVM. The system then constructs a new SVM to replace the current SVM based on the SVMs built for each subset of violating data vectors.
|
