We are working on prediction one company production estimation and the main field of works is like stock market prediction(Time series analysis and process data).

So I have some comment on using finite impulse response (FIR) method for feature selection and not using the simple regression method for This purpose.

The reason for this suggestion is considering more the 1 order of X in FIR method vs the simple regression, as you can see below:

enter image description here

So I have googled the below term and not found any post or other resource to read more about it.

Simple regression vs finite impulse response (FIR)

If possible, I have asked here to have your comment to find more information about this feature selection method.




You must log in to answer this question.

Browse other questions tagged .