Can I use ArcPro to do machine learning on numeric (point) data?

[TedR]

Hi - I am looking at how best to do machine learning on a set of spatial data (latitude, longitude, variable 1 (numeric), variable 2 (numeric), variable 3 (classified nominal - i.e. can be one of four values).  Can ArcGIS be used for this?  Is there a recommended tutorial or example of this?   (I see where ArcPro can do machine learning on raster data but that is not my need.)

My goal is to predict variable 1 based on all the rest of the inputs described above.  Thanks!

[Robert_LeClair]

There are several free tutorials on machine learning using ArcgIS Pro - click here to see the many self paced modules.  To learn how to use machine learning in ArcGIS Pro, the help documentation  is really good too!

[TedR]

Thanks Robert.  The documentation I have read all pertains to machine learning on raster/image data.  I am not interested in that.  Instead, I need to do machine learning on spatially referenced numeric table data.  So my question has to do with whether or not ArcPro can machine learn on spatially referenced numeric table data and, if so, where I can go to learn how to apply this to my dataset.   Have you or anyone else heard of doing this - it seems like this would be a common request.  Maybe I'm not looking in the right place.  

[Robert_LeClair · Accepted solution]

You're welcome Ted.  So in a quick, cursory review of "machine learning" and "tabular data", it does appear it's possible via the ArcGIS API for Python Developers page - Machine learning and deep learning on tabular data | ArcGIS API for Python

Caveat here - I'm not a Python user nor a Deep Learning expert but based upon my reading of this page, it does seem possible.

[TedR]

Thanks again.  I'll have a look!

[Robert_LeClair]

Certainly!  Let me know what you discover!