Answered
Combining two deep neural networks to train simultaneously
Hello, You can do this in Simulink - see the following examples for reference. https://www.mathworks.com/help/reinforcement-l...

8 månader ago | 1

| accepted

Answered
DQN learns at first but then worsens.
To confirm that this is an exploration issue, can you try setting the EpsilonMin param to a high value? e.g. 0.99. If after doin...

8 månader ago | 0

Answered
How to resume train a trained agent?about Q learning agents.
Hello, To see how to iew the table values, take a look at the answer here. Also, you don't have to do anything specific to con...

8 månader ago | 0

| accepted

Answered
Reinforcement learning action getting saturated at one range of values
Your scaling layer is not set up correctly. You want to scale to (upper limit-lower limit) and then shift accordingly. scaling...

8 månader ago | 0

| accepted

Answered
How can I provide constraints to the actions provided by the Reinforcement Learning Agent?
Hard constraints are not typically supported during training in RL. You can specify limits/constraints as you mention above, but...

8 månader ago | 0

| accepted

Answered
Exporting data only works as pdf. Axis labels are getting small and unreadable
You cannot save as .fig from the episode manager plot. If you have the training data though (it's good practice to save this dat...

8 månader ago | 1

| accepted

Answered
Reinforcement Learning multiple agent validation: Can I have a Simulink model host TWO agents and test them
That should be possible. Did you follow the multi-agent examples? Since the agents are trained already you may want to check the...

8 månader ago | 0

| accepted

Answered
Do the actorNet and criticNet share the parameter if the layers have the same name?
No, each network has its own parameters. Shared layers are not supported out of the box, you would have to implement custom trai...

8 månader ago | 0

| accepted

Answered
Any RL Toolbox A3C example?
Hello, To get an idea of what an actor/critic architecture may look like, you can use the 'default agent' feature that creates ...

8 månader ago | 0

| accepted

Answered
After training my DDPG RL agent and saving it, unexpected simulation output
See answer here

8 månader ago | 0

| accepted

Answered
Saved agent always gives constant output no matter how or how much I train it
The problem formulation is not correct. I suspect that even during training, you are seeing a lot of bang bang actions. The bigg...

8 månader ago | 1

| accepted

Answered
How can I create a Reinforcement Learning Agent representation based on Recurrent neural network (RNN, LSTM, among others)
Hello, Which release are you using? R2020a and R2020b support LSTM policies for PPO and DQN agents. Starting in R2021a you can ...

8 månader ago | 2

| accepted

Answered
Procedure to link state path and action path in a DQL critic reinforcement learning agent?
Hello, Some comments on the points you raise above: 1.There are two ways to create the critic network for DQN as you probabl...

8 månader ago | 0

| accepted

Answered
Reinforcement learning DDPG Agent semi active control issue
Hello, This is very open-ended so there could be a lot of ways to improve your setup. My guess is that the issue is very releva...

8 månader ago | 1

| accepted

Answered
Save listener Callback in eps format or any high resolution format
Hello, If you are using R2020b, you can use help rlPlotTrainingResults to recreate the Episode manager plot and save it as y...

8 månader ago | 0

| accepted

Answered
Input normalization using a reinforcement learning DQN agent
Hello, Normalization through the input layers is not supported for RL training. As a workaround, you can scale the observations...

8 månader ago | 1

| accepted

Answered
Export Q-Table from rlAgent
Here is an example load('basicGWQAgent.mat','qAgent') critic = getCritic(qAgent); tableObj = getModel(critic); table = table...

8 månader ago | 1

| accepted

Answered
Replace PI Controller with RL Agent for simple Transfer Function
Please see answer here: https://www.mathworks.com/matlabcentral/answers/779177-ddpg-agent-isn-t-learning-reward-0-for-every-epi...

9 månader ago | 1

| accepted

Answered
DDPG Agent isn't learning (reward 0 for every episode)
The reason why you see 0 rewards is because thw IsDone flag (which is used to terminate episodes early) is immediately set to tr...

9 månader ago | 1

| accepted

Answered
Transient value problem of the variable in reward function of reinforcement learning
You can put the agent block under a triggered subsystem and set it to begin training after 0.06 seconds

9 månader ago | 0

| accepted

Answered
Agent is suddently doing random actions and training diverge
This is normal behavior - one common misconception is that once the reward starts going up, it will remain up. This is not true ...

9 månader ago | 1

| accepted

Answered
Reinforcement Learning does not show that training occurs?
Thanks for the info. I think this is a scaling issue with the plot. The Episode Manager has this option where you can uncheck "Q...

9 månader ago | 0

Answered
Reinforcement Learning Onramp Issue
Please take a look at this answer.

9 månader ago | 0

Answered
Creating Q-table
Did you take a look at this example? It seems to solve a similar problem. If you want to use the provided API to create a custo...

9 månader ago | 0

Answered
Read data from csv file into a reward function for Reinforcement Learning
It seems like you were trying to read the file from within the MATLAB Fcn block (this block assumes that anything you write in i...

9 månader ago | 0

| accepted

Answered
Reinforcement learning : How to define custom environment with multiple image based observations
For grayscale images, take a look at this example. For rgb, maybe the following would work ObservationInfo = rlNumericSpec([320...

9 månader ago | 0

Answered
How to avoid repeated actions and to manually end episode for a DQN agent?
From what you are saying, it seems that training has not converged yet. During training, the agent may every now and then behave...

9 månader ago | 0

Answered
Set gpu option for rlPPOAgent actor
What you have specified is sufficient for the critic. If you do the same for the actor you are all set - there is no additional ...

9 månader ago | 0

Answered
Reward in training manager higher than should be
Cannot be sure about the error, but it seems somewhere in your setup you are currently changing changing the number of parameter...

9 månader ago | 0

Answered
Visualize Progress in Reinforcement Learning Toolbox
This is not possible out of the box, but you could implement something like this by setting a counter and saving the current ve...

9 månader ago | 0

| accepted

Load more