Chih-Jen Lin

I think due to numerical inaccuracy, both c-svc and nu-svc may be slightly affected by the order of data. However, the resulting model (or test accuracy) should be similar Alexey...

csvc and nusvc are equivalent. Are you using the corresponding parameters? Alexey Rozhkov writes: > In this case it looks like a bug - accuracy variation is too > high...

In 3.17 there is a change: 3.17: 2013/04/01 svm.cpp: if class labels are 1 and -1, ensure labels[0] = 1 and labels[1] = -1 this might cause model incompability. Arun...

will be added in the next release: diff --git a/README b/README index 1db0e8c..7c1f889 100644 --- a/README +++ b/README @@ -42,8 +42,9 @@ More information about parameter selection can be found...

C-svc and nu-svc are equivalent. So when you train nu-svc we print out the corresponding C marcwell writes: > When I train a nu-SVC in Octave with the command >...

What's the reference of one-class svm with negative samples? BTW the optimization algorithm in libsvm for one-class svm does assume y=1.. So you cannot just change y to have 1...

Your modification doesn't solve their opt problem. See solve_one_class() in svm.cpp: ``` for(i=0;inu * prob->l - n; ``` The init sol is under the assumption of e^Talpha = a positive...

added in libsvm 3.23 Dhruv Madeka writes: > There doesn't seem to be a wrapper in Python for the function > > — > You are receiving this because you...

See python/README On 2018-08-02 22:18, Akbar Gumbira wrote: > @cjlin1 [1] hi, where can I find it? is it possible to use it using > the same scaling parameter file?...

Do you mean the weight vector? If so, yes they can be larger than 1 naarkhoo writes: > I wonder, if coefficients can be larger than 1 even on the...