Thanks Ronnie,

> I think you're looking for segmented regression models. 

Those papers and your suggested search terms were very helpful.

I’m still making my way through them but a first glance they (and others I 
found in my searches) are not taking the time sequencing into account and the 
Pattern Recognition in Time Series articles I found all worked with a single 
output variable.  Its slow going for me as many years in commercial IT have 
atrophied my mathematical and statistical skills.

I think I’m just going to have to bite the bullet and work up some code that 
heuristically builds the partitions (maybe manually guided) using scipy-learn 
and see if that helps me understand the structure of the data better.

Thanks for taking the time to answer a somewhat off-topic question.

Cheers, 
Dieter.
------------------------------------------------------------------------------
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

Reply via email to