Optimization of the parameters for a production System

alexanderthelin

I have already searched the community and Youtube, i can not believe that there isn't already a answer to my problem, but i can not find the answer.

I Simulated data in a Material flow Simulation. I already build a Neural Network out of that data, which is pretty accurate. Now i want to find the optimal parameters for this Production System. Can someone show or link me an example that is similar?
I want to use the already build neural network and an evolutionary algorithm to optimize it.
When i search for something like optimization i can only find examples where a process in Rapidminer is optimized. But i want to find the best parameter combination for the input data.

Thanks for you help
Alex

IngoRM

Hi,

I am not really sure if you are actually asking about the optimization of the model parameters.  If you are, the answer from @kayman has all the right hints and will get you going.

You may also have been asking about figuring out the optimal inputs to a model for getting a desired outcome (also known as prescriptive analytics).  That is possible as well.  The key operators for this are called Model Simulator and Prescriptive Analytics. 

There is a video of the simulator and the optimization (which by the way uses indeed evolutionary optimization) here:

https://academy.rapidminer.com/learn/video/auto-model-classification

It starts around minute 6:40 and shows the stuff as part of the Auto Model results.  But you can create the same things as part of the process (Model Simulator) and even produce the optimal settings for further use in the process (Prescriptive Analytics).

Hope this helps,
Ingo