Support vector machine output parameters for one-vs-one classification?

Question

Imran Riaz 2022-9-30

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1815035-support-vector-machine-output-parameters-for-one-vs-one-classification

回答： Kartikay Sapra 2022-10-3

I have done training and testing stage. I used SVM for my 90 different classes and performed one-vs-one and one-vs-all classification. I got the results of pecision, recal and F1 score for both OvO and OvA(also confusion matrix). Now how I have to put these results in my paper? Which is the best way? Either i need curves or plot some graphs.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Kartikay Sapra 2022-10-3

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1815035-support-vector-machine-output-parameters-for-one-vs-one-classification#answer_1065520

在 MATLAB Online 中打开

You can use the following in-built functions to compute and plot Confusion Matrix:

%Assuming y is the true label vector, and pred_y is the predicted label vector
>> confusion_mat = confusionmat(y, pred_y)
>> confusionchart(confusion_mat) %to plot the confusion matrix

Refer to:

https://in.mathworks.com/help/stats/confusionmat.html

For other useful plots:

https://in.mathworks.com/help/stats/perfcurve.html