"High dimensional regression - the best (and fastest) approach with RM?"

brett_800

I'm looking to model a regression target (ie: a continuous target). I have data sets that typically have more variables than observations.

What RM processes are best for feature selection/dimension reduction, and then modelling targets of this nature?

venkatesh

You can attempt stochastic gradient descent based linear regression using the Optimization extension. You can use this for high dimensional and huge datasets

The extension is available athttps://bitbucket.org/venkatesh20/optimization-extension

a video tutorial on this extension is available athttps://www.youtube.com/watch?v=nECDYNT5owg