0
$\begingroup$

I have two machines, Machine A and Machine B collecting Time Series data. The first machine runs every day and collects 5 features, the second runs every Friday and collects 10 features. Trying to apply this to a linear regression model, but the overlap and difference in feature count is making it difficult. What I could do is append 5 zeros to the daily features to make it work. But this seems a little hacky. What's the best practice here?

$\endgroup$

1 Answer 1

0
$\begingroup$

Appending 5 zeros is probably not a good approach. Interpolation (or some other data imputation technique) might be better here.

An alternative is, to treat the data from the two machines separately (but without more information about the problem, it is difficult to jedge if this is feasible or not).

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.