How can I split data train, validation using for classificaiton learner

Question

beyoglu 2023-6-5

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1978414-how-can-i-split-data-train-validation-using-for-classificaiton-learner

评论： Govind KM ，2023-6-6

Hello, is there a way to select the train and validation data separately from the separate data in the workspace for classification learner? Classification Learner interface does it by splitting the train data into %train and %test as a deafult. I need this to take advantage of the classification learner.

Kind regards

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

Govind KM 2023-6-5

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1978414-how-can-i-split-data-train-validation-using-for-classificaiton-learner#answer_1250539

You can set the percentage of data to be reserved for testing or even import a test data set through the Test section on the Classification Learner tab. You can refer to these documentation sites for more information :

https://in.mathworks.com/help/stats/select-data-and-validation-for-classification-problem.html (Section : Reserve Data for Testing)

https://in.mathworks.com/help/stats/assess-classifier-performance.html#mw_bce8ad28-f1d5-48fa-be79-9c272149a135 (Section: Evaluate Test Set Model Performance)

2 个评论
显示无隐藏无

beyoglu 2023-6-5

Let's assume that we have data from three separate datasets, train validation and test. We will evaluate their performance. In the link, the random part of the data as train and test is explained.

best regards

Govind KM 2023-6-6

While creating a new session, you can select holdout validation and select a percentage of data to be used as a validation set. This can help you if you want to set a specific number of samples for validation.

I don't think you can specify the validation data separately as of now.

请先登录，再进行评论。