SVM in classification.
SVM in classification.Try to find the hyperplane as minimize
(where if , else )
(in fact, we use a kernel instead of a dot product)
The coefficients are given in C_ when you call the constructor. If this one is NULL, then the value given by the "C" option is used for all . (The size of C_ must be data->n_real_examples)
Options:
"C" real trade off between the weight decay and the error [100] "cache size" real cache size (in Mo) [50]
Alphabetic index HTML hierarchy of classes or Java