How to get the mathematical equation of a model trained using Regression learner toolbox

I have created an SVM model using the regression learner. I have exported the trained model as well. How do I get the mathematical equation behind the model, that has all the coefficients and parameters that have been computed in the process of training?

[Ive J]

Not possible. The best you can get is the model object itself, and use predict method to predict over a new set of predictors.

[Nathan Williams]

That is ridiculous. How is running a .m function preferable to executing a mathematical equation?

Answer 1

[Ravi]

Hi Sirish Selvam,

I understand that you want to find the mathematical equation of the trained SVM model. Unfortunately, the mathematical equation of SVM is quite complex unlike the linear regression models.

There is no direct way to obtain the mathematical equation of the trained SVM, but you can still obtain the parameters, coefficients and bias term of the support vector machine model trained by accessing the model properties.

Let us assume the model trained is “svm_model”. We can access the support vectors, coefficients, and bias term as follows.

% Extract support vectors and coefficients 
supportVectors = svm_model.SupportVectors; 
coefficients = svm_model.Alpha; 

% Extract bias term (if present) 
bias = svm_model.Bias; 

To learn more about the SVM model and its properties, please refer to the following link.

fitrsvm - https://www.mathworks.com/help/stats/fitrsvm.html

I hope this solution resolves the issue you are facing.

Thanks,

Ravi Chandra