Even if it was for multivariate time series, the model would first need to infer what machine are we talking about, then its working conditions, and only then make a reasonable forecast based on an hypothesis of its dynamics. I don’t know, seems pretty hard.
Comments
Even if it was for multivariate time series, the model would first need to infer what machine are we talking about, then its working conditions, and only then make a reasonable forecast based on an hypothesis of its dynamics. I don’t know, seems pretty hard.
Indeed. An issue I ran into over and over while doing research for semiconductor manufacturing.
My complaint was more illustrative than earnest.