image

🎉 🎉 RAPIDMINER 9.10 IS OUT!!! 🎉🎉

Download the latest version helping analytics teams accelerate time-to-value for streaming and IIOT use cases.

CLICK HERE TO DOWNLOAD

How to pass winning Hidden Layer Sizes for DL to another DL learner?

rapid47661913rapid47661913 Member Posts: 3 Contributor I
edited November 2018 in Help

I’m trying to create a genetic deep learning system of sorts, but the passing of Hidden Layer Sizes won’t work. Here’s the concept:

  • Use an Optimize Parameters to pick a winning Deep Learning system (run 10 epochs).
  • Pass the Hidden Layer Sizes to another DL system.
  • Generate a new DL model using the HLS and 100 epochs, etc.

You cant directly specify Hidden Layers in an Optimize Parameters with DL, so you have to pass the values as macros. The problem is that, for some reason, I can’t capture the winning Sizes and store them either as parameters or a macro that I can then pass on.

Is this not a feasible idea? What approach should I use? I’d show you what I’ve done, but I’ve tried so many combinations that everything is a total jumble right now and would probably only confuse the situation.

 

Thanks in advance.

Sign In or Register to comment.