How to develop regression function (Mathmatical form) from trained model using regression learner?

2 次查看（过去 30 天）

Manoj kumar Beriya 2021-2-15

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/746027-how-to-develop-regression-function-mathmatical-form-from-trained-model-using-regression-learner

评论： Manoj kumar Beriya 2021-2-18

How to extract regression function in term of mathmatical form (or tool or excel sheet) from trained model using regression learner (Support Vector Regression).

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。

采纳的回答

Aditya Patil 2021-2-18

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/746027-how-to-develop-regression-function-mathmatical-form-from-trained-model-using-regression-learner#answer_626402

在 MATLAB Online 中打开

In case of support vector machines(both for classification and regression), the score function/regression function might not be in terms of input predictors, hence it might not be possible to get the regression function. The coefficients can be obtained using

trainedModel.RegressionSVM

As a workaround, you can use other regression techniques such as linear regression, which derive output in terms of predictors.

5 个评论
显示 3更早的评论隐藏 3更早的评论

Aditya Patil 2021-2-18

编辑：Aditya Patil 2021-2-18

You can export the model, and create a function that takes in a value, calls the predict function, and returns the output. Then call this function from excel using matlabfcn.

If you need further help, feel free to create a new question, as that would reach out to more people.

Manoj kumar Beriya 2021-2-18