sgd_svm¶

besca.tl.auto_annot.sgd_svm(train, y_train)[source]¶

fits linear svm on training dataset using stochastic gradient descent

Parameters:

train (pd.DataFrame) – adata_train.X but scaled and as a dataframe
y_train (pd.DataFrame) – one dimensional dataframe containing class label

Returns:

trained svm classifier

Return type:

sklearn.calibration.CalibratedClassifierCV