Hello Community, I came across an AutoML platform that uses mathematical tangential modeling techniques for time series forecast modeling . Any comparative study out there that evaluates this approach to the more main stream DL and RNN in terms of speed, infrastructure requirements and dealing with categorical data.
This is exactly what we have put into our time series platform. Our time series product is not available in the trial. Feel free to contact DataRobot if you would be interested in evaluating/purchasing it for your enterprise.