Hi all, I have successfully implemented the reinforcement learning with the guide given. However, I am curious as to how to implement in the following scenario when working with lists. Namely, there are a few key differences:
1. Instead of discrete numbers as the type, now we have very different types with strings, ie. "A102", "B103", etc. I understand that this could possibly be done through the use of options instead of discrete. Just wanted to make sure that this is possible.
2. Instead of optimizing for one processor, now I am trying to optimize for 3.
3. Instead of pulling the best item that matches the parameters in the example, I am now pulling from lists instead. There are also few key criteria in the pulling: Processor = 1, Processor = 2, etc. These are to be kept and should not be changed.
Extra Separate Scenario:
Additionally, would it be possible for RL to be working in Flexscript? I am building one that simulates queue shortages and aims to improve throughput. In the following scenario, I have 3 types of items, A B and C with 10 queues that can only store one item each.
Currently, Flexscript simply allocates whichever queue is available. Type C is stuck if Type A and B takes them all.
Items A and B can be stored in all 10 queues but item C can only be stored in 3 specific queues. At some point in time, I need to make the decision whether to also give up item C's queue to allow A and B to take them. I suppose this could be an implementation of RL but I'm having trouble understanding how I can incorporate it within Flexscript (I'm currently writing the queue assignment logic in Flexscript since the actual logic is more complex).
