reduce_class_svm — Approximate a trained support vector machine by a reduced support vector machine for faster classification.
As described in create_class_svm, the classification time of a SVM depends on the number of kernel evaluations between the support vectors and the feature vectors. While the length of the data vectors can be reduced in a preprocessing step like 'principal_components' or 'canonical_variates' (see create_class_svm for details), the number of resulting SV depends on the complexity of the classification problem. The number of SVs is determined during training. To further reduce classification time, the number of SVs can be reduced by approximating the original separating hyperplane with fewer SVs than originally required. For this purpose, a copy of the original SVM provided by SVMHandle is created and returned in SVMHandleReduced. This new SVM has the same parametrization as the original SVM, but a different SV expansion. The training samples that are included in SVMHandle are not copied. The original SVM is not modified by reduce_class_svm.
The reduction method is selected with Method. Currently, only a bottom up approach is supported, which iteratively merges SVs. The algorithm stops if either the minimum number of SVs is reached (MinRemainingSV) or if the accumulated maximum error exceeds the threshold MaxError. Note that the approximation reduces the complexity of the hyperplane and thereby leads to a deteriorated classification rate. A common approch is therefore to start from a small MaxError e.g., 0.001, and to increase its value step by step. To control the reduction ratio, at each step the number of remaining SVs is determined with get_support_vector_num_class_svm and the classification rate is checked on a separate test data set with classify_class_svm.
Original SVM handle.
Type of postprocessing to reduce number of SV.
Default value: 'bottom_up'
List of values: 'bottom_up'
Minimum number of remaining SVs.
Default value: 2
Suggested values: 2, 3, 4, 5, 7, 10, 15, 20, 30, 50
Restriction: MinRemainingSV >= 2
Maximum allowed error of reduction.
Default value: 0.001
Suggested values: 0.0001, 0.0002, 0.0005, 0.001, 0.002, 0.005, 0.01, 0.02, 0.05
Restriction: MaxError > 0.0
SVMHandle of reduced SVM.
* Train an SVM create_class_svm (NumFeatures, 'rbf', 0.01, 0.01, NumClasses,\ 'one-versus-all', 'normalization', NumFeatures,\ SVMHandle) read_samples_class_svm (SVMHandle, 'samples.mtf') train_class_svm (SVMHandle, 0.001, 'default') * Create a reduced SVM reduce_class_svm (SVMHandle, 'bottom_up', 2, 0.01, SVMHandleReduced) write_class_svm (SVMHandleReduced, 'classifier.svm') clear_class_svm (SVMHandleReduced) clear_class_svm (SVMHandle)
If the parameters are valid the operator train_class_svm returns the value 2 (H_MSG_TRUE). If necessary, an exception is raised.
classify_class_svm, write_class_svm, get_support_vector_num_class_svm