Is it possible to calculate the importance of the input variable using weights, beta and bias calculated by SVM and NN?

Hello.

Is it possible to automatically calculate the importance of an input variable using weights, beta, and bias calculated by SVM and NN? Or is there a separate calculation formula?

thank you

Answer 1

[Aditya Patil]

In case of SVM, it's generally not possible to calculate feature importance. You can find more details regarding this here. In case of neural networks, it's not straightforward, and is an open research topic.

You can use the various dimensionality reduction techniques available in MATLAB instead.