What is the predicted score and how can I use it for the distance of the datapoint to the hyperplane?

1 回表示 (過去 30 日間)
I am coding a support vector machine. When I try this code
[label,score] = predict(Model,X_test_best_feature);
then I get something like this i.e. two columns for the score which confuses me.
table(y_test, label, score)
y_test label score
______ _____ ____________________
2 2 -0.99984 0.99984
2 2 -0.99956 0.99956
2 2 -0.99983 0.99983
2 2 -0.99956 0.99956
2 2 -0.99969 0.99969
And how can I use the score to compute the distance of each datapoint to the hyperplane?
  1 件のコメント
Krishna Bindumadhavan
Krishna Bindumadhavan 2018 年 7 月 9 日
Did you already have a look at the documentation here: https://www.mathworks.com/help/stats/compactclassificationsvm.predict.html#d119e588869 ?
There is information regarding score both at "Output arguments" and at the end under "More About".
Specifically, the first column contains negative class scores and the second column contains positive class scores.

サインインしてコメントする。

回答 (0 件)

タグ

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by