使用一类svm-python进行新颖性检测

Question

I'm in the process of novelty detection using machine-learning. I have tried using one-class svm in scikit learn .

from sklearn import svm

train_data = [[0, 0, 0, 0, 0, 1, 0, 0], [0, 1, 0, 0, 0, 1, 0, 0], [0, 1, 0, 0, 0, 0, 0, 0], [0, 1, 0, 0, 0, 0, 0, 0], [0, 1, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 1], [0, 3, 0, 0, 0, 1, 0, 0], [0, 11, 0, 0, 0, 0, 0, 0], [0, 1, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 4]]
test_data = [[0, 0, 0, 0, 0, 1, 0, 0], [0, 1, 0, 0, 0, 1, 0, 0]]


clf = svm.OneClassSVM(nu=0.1, kernel="rbf", gamma=0.1)
clf.fit(train_data)

pred_test = clf.predict(test_data)

I'm new to this area and I want to know how can I say there is novelty in my test data?

Answer 1

The inliers are labeled 1, and the outliers (ie, the novelties in your case) are labeled -1 (as the result of the predict function).

Please notice that the current documentation incorrectly states that the outliers are labeled 1 & inliers are labeled 0. Please check out the latest updates on github repo for the correct information.

Answer 2

check = clf.predict(test_data)

if check = 1 then not anomaly and

if check = -1 then it an anomaly ie data is outlier