I would like to know, machine learning about optimization

Question

question

mark zhen asked Aug 15 2022 at 4:56 PM Jeanette F commented Sep 06 2022 at 4:00 PM

I would like to know, machine learning about optimization

I would like to know, machine learning about optimization,

Is there any information on how I can improve the model made by the teaching guide?

(The way I want it to work is as follows, I have 5 products today with a random parameter, and I want to use reinforcement learning to train an optimized production schedule.)

Software Version:

FlexSim 22.0.0

python production schedule

· 1

Answer 1 · 2022-08-15T21:24:31Z

Kavika F answered Aug 15 2022 at 9:24 PM Jason Lightfoot commented Aug 31 2022 at 1:48 PM

Here are the docs on Reinforcement Learning Training, Concepts of Reinforcement Learning, and How to Use a Trained Model. This will point you in the right direction and hopefully provide enough detail you're looking for.

· 16

mark zhen commented · Aug 16 2022 at 4:44 AM

I have already practiced the above content, but there is still a gap between what I expected. What I need is something that can be collected into a table. Let me know the order of production scheduling.

0 ·

Felix Möhlmann mark zhen commented · Aug 16 2022 at 8:25 AM

Once you have trained the algorithm you can run the model while the algorithm is making decision during the run.

You could then record those decisions or their effect on the model by listening to various events and writing data to a global table.

For example to record the actions the algorithm is taking, you can listen to the RL-event "OnRequestAction" in an event-triggered source (use the sampling tool on the RL entry in the toolbox to select the event). Then have the token wait in a Delay activity for a tiny amount of time (so the request is processed first) and then log the value of the action parameters in a global table.

Similarly, you could log the item types that are processed by listening to some trigger of the processor (OnEntry, OnProcessFinish, ...).

What data you should collect depends on what exactly your desired output is.

0 ·

1660638184086.png (47.1 KiB)

mark zhen Felix Möhlmann commented · Aug 16 2022 at 8:28 AM

但我想要做的事情是類似最佳化生產排程的計算，我該如何利用強化學習來達成這件事。

But what I want to do is something like optimizing production scheduling, and how can I use reinforcement learning to do that?

0 ·

Show more comments

question