How to preprocess, geoencode and add multivariate regression-based time series dataset for deep learning modelling in arcgis pro?

11-04-2022 09:11 AM
New Contributor II

I have spent several nights on this particular issue. I need help.

I have a regression-based modelling task (i.e. I wish to predict continuous values). I have a multivariate weather dataset (precipitation, solar radiation, wind speed) collected at three different locations daily for two years. 

My approach is to create a multidimensional space-time-cube typical in cluster analysis, such that the three variables will form an overlay of cubes at each gauge location temporally. 

How do I preprocess, geoencode, and add my dataset to the Level 4 administrative shapefile that I have got?

The goal is to predict one of the variables given the other two variables spatio-temporally i.e. at given time steps and at each gauge location.

It's a serious concern to me. Thank you.

0 Kudos
0 Replies