CODEWITHSUNDEEP
machine-learning
Jorgemar
Jorgemar

Reputation: 83

Getting a mapping function between two time series

I wonder if there is a method to get the function that transforms one time series in another.

For example, for the time series t1=[1 0 1 0 2] and t2=[10 0 10 0 20], the mapping function would be t2 = t1*10.

This is an easy case that can be even manually computed, but I need to get this function for much more complex cases.

Do you know a way to do that?

Thank you in advance

Upvotes: 1

Views: 305

Answers (1)

Ben Allison
Ben Allison

Reputation: 7394

Are the number of elements the same in the two series? If so, why not just build a regression model with xs as elements from the first time series and ys from the other. Unless the mapping is sequential in some way (i.e. x_i -> y_i depends on i), this should work pretty well. If there is temporal dependence then you could build a more complex model where i was a predictor in the regression.

If you say what language you're using, I (or someone else) should be able to suggest a library, or post example code.

Upvotes: 1

Related Questions