Hi,<br>I have a couple of queries about multiclass classification.<br>I have 3 classes and I want to classify them using SVM with Searchlight. <br><br>Code snippet:<br>############################<br>dataset = dataset.selectSamples(<br>
N.array([l in [1, 2, 3] for l in dataset.labels],<br> dtype='bool'))<br><br>mlclf = MulticlassClassifier(clf=LinearCSVMC(),bclf_type="1-vs-1")<br><br>cv = CrossValidatedTransferError(<br>
TransferError(mlclf),<br> CustomSplitter([([0,1],[2,3,4,5,6,7,8,9,10,11])],<br> attr='chunks')<br> )<br><br>sl = Searchlight(cv, radius=3)<br>
<br># run searchlight on example dataset and retrieve error map<br>sl_map1 = sl(dataset)<br><br>################################<br>Queries:<br><br>1) I am getting the following warning. What does it mean?<br><br>WARNING: Boosted classifier MulticlassClassifier(clfs=[BinaryClassifier(clf=LinearCSVMC(kernel_type='linear', svm_impl='C_SVC'), poslabels=[1.0], neglabels=[2.0]),...], combiner=MaximalVote(), bclf_type='1-vs-1', clf=LinearCSVMC(kernel_type='linear', svm_impl='C_SVC')) has 'values' state enabled, but combiner has it active, thus no values could be provided directly, access .clfs<br>
<br><br>2) If I don't use Multiclass, there is no problem. Should I use Multiclass at all?<br><br>Thanks,<br>Geethmala<br>