![]() ![]() In one-class SVM, parameter nu should be an upper bound of the training error rate.I would like to solve L2-loss SVM (i.e., error term is quadratic).What method does libsvm use for multi-class SVM ? Why don't you use the "1-against-the rest" method?.Any example? In svm.cpp, there are two subroutines for kernel evaluations: k_function() and kernel_function(). How do I disable screen output of svm-train?.large -m) on a linux machine, why sometimes I get "segmentation fault ?" On 32-bit machines, if I use a large cache (i.e.How do I get the distance between a point and the hyperplane?.The program keeps running (with output, i.e.The program keeps running (without showing any output).What is the difference between nu-SVC and C-SVC?.Does it make a big difference if I scale each attribute to instead of ?.On windows sometimes svm-scale.exe generates some non-ASCII data not good for training/prediction?.Should I scale training and testing data in a similar way?.The number of free support vectors is large.Does libsvm have special treatments for linear SVM?.Should I use float or double to store numbers in the cache ?.Can you explain more about the model file?.The output of training C-SVM is like the following.How to convert other data formats to LIBSVM format?.Is there a program to check if my data are in the correct format?.Why sometimes the last line of my data is not read by svm-train?. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |