Emmanouil Tzorakoleftherakis

Last seen: Today

MathWorks

98 total contributions since 2018

Emmanouil Tzorakoleftherakis's Badges

Answered
Easy way to evaluate / compare the performance of RL algorithm
Why not use a MATLAB Fcn block and implement the dummy agent in there? If you want random/constant actions should be just one li...

4 days ago | 1

Answered
Is it possible to train LSTM Network without a Dataset?
In the paper they mention "Although a readily available dataset is required to train an LSTM network, we devised an efficient wa...

11 days ago | 1

| accepted

Answered
Reinforcement learning: "NextObs" vs. "LoggedState" in step function
Actually, NextObs is the important thing here. It represents the value of your states after you apply current action and integra...

11 days ago | 0

Answered
What's the purpose of adding a transfer function after a Integrator block?
Hello It is likely to filter high-frequency content. Hope that helps.

12 days ago | 0

Answered
PPO agent with continuous action example
Hello, If you want to use PPO, i.e. a stochastic actor with continuous action space, you can follow the structure shown here.

16 days ago | 0

Answered
Environment for Reinforcement Learning Project
Hello, We are working on providing an interface between OpenAI Gym and Reinforcement Learning Toolbox but this will take some m...

17 days ago | 0

Answered
How do I properly substitute rlRepresentation with rlValueRepresentation, rlQValueRepresentation, rlDeterministicActorRepresentation, and rlStochasticActorRepresentation?
It would be helpful if you pasted the exact MATLAB code you are typing to see what the problem is. I suspect you simply changed ...

22 days ago | 0

Answered
Deep Q Learning - define an adaptive critic learning rate?
Hi Niklas, I believe this is currently not supported. This is an interesting usecase though - I will inform the development tea...

23 days ago | 0

| accepted

Answered
Build Environment reinforcement learning
Hello, For Simulink environments, the following page should be helpful: https://www.mathworks.com/help/reinforcement-learning/...

23 days ago | 0

Answered
Using Reinforcement Learning algorithm to optimize parameter(s) of a controller
Hi Hazwan, The main difference between using RL for control vs parameter tuning is that in the first case the policy will direc...

23 days ago | 1

| accepted

Answered
Initializing pimp-controller failed: Error binding to tcp://*: no free port in range 9620-9620
Hello, I would contact technical support for this, and show them how to reproduce the error. If the issue is in the communicati...

23 days ago | 0

Answered
Can LoggedSignal in provided Link contain more than just the state?
LoggedSignals is not tied to the state or the observations, so you should be able to store whatever makes sense to you in that v...

25 days ago | 0

| accepted

Answered
Using getValue in matlab fcn block in simulink
Hi Sam, Before R2020a, the easiest way to bring the critic in Simulink without using the Agent block is to call generatePolicy...

25 days ago | 1

Answered
Multi action agent programming in reinforcement learning
This example shows how to create an environment with multiple discrete actions. Hope that helps

25 days ago | 0

Answered
Incorporate Time into Reinforcement Learning Environment
Time would be another parameter of your environment. Interactions between the agent and environment happen at discrete time step...

25 days ago | 1

| accepted

Answered
How to view the output of rlNumericSpec?
Hi Jacob, I think what you want to do is take the output of the agent and do the transformation you mention (not the output of ...

25 days ago | 0

Answered
Create and Train DQN Agent with just a State Path and Not Action Path
Hello, This page shows how this can be done in 20a. We will have examples that show this workflow in the next release. Hope th...

1 month ago | 1

| accepted

Answered
To choose an action, is it correct to compute the value of successor state or do we need to compute value of states in the entire path till end state?
Hi Gowri, Using the Q value for a state+action pair encodes all the information till 'the end of the path' weighted by a discou...

1 month ago | 1

| accepted

Answered
Agent repeats same sequence of actions each episode
Hi Braydon, I am not really sure why you are only looking at the first two episodes. RL can take thousands of episodes to conve...

1 month ago | 0

| accepted

Answered
How to create an custom Reinforcement Learning Environment + DDPG agent
Hi Kai, What the very first error is telling you is that there is an issue with the dimensions of either your observation, rewa...

1 month ago | 1

| accepted

Answered
Reinforcement Learning Sample Time
Hi Braydon, The agent sample time effectively determines how often the agent will output a decision/action. Think of it as the...

1 month ago | 0

Answered
2 out of 7 Observations Defined in MATLAB DDPG Reinforcement Learning Environment. Are the rest given random values?
Hello, I am assuming you are referring to the initialization of x and y inside the "flyingRobotResetFcn" function. Basically, i...

1 month ago | 0

| accepted

Answered
RL Agent does not learn
Hello, It is really hard to say just by looking at the training plot. The first thing I would try is 1) a different agent (mayb...

1 month ago | 0

| accepted

Answered
Regarding Reinforcement Learning with GridWorld
Hello, You would need to create a custom Grid world for this. Have a look at this link, and particularly the 'Actions' property...

1 month ago | 0

| accepted

Answered
Why is the DDPG episode rewards never change during the whole training process?
Looks like the scale between Q0 and episode reward is very different. Try unchecking "Show Episode Q0" to see of the episode rew...

2 months ago | 0

| accepted

Answered
Reinforcement Learning -- Rocket Lander
Hi Averill, Here is a version that converges in ~18-20k episodes - thank you for pointing out that this example was not converg...

3 months ago | 0

| accepted

Answered
QTable reset when using train
If you stop training, you should be able to continue from where you left off. I called 'train' on the basic grid world example a...

3 months ago | 0

| accepted

Answered
How to test critic network in DDPG agent?
You can use getValue for inference on critic networks.

3 months ago | 1

| accepted

Answered
Questions about Reinforcement Learning
Hi Averill, Can you please let me know which examples do not converge? I will share this information with the development team....

3 months ago | 0

Answered
Error while designing MPC controller where output of vehicle body block is using as measured output in MPC
Seems the error is self-explanatory - have you checked the dimensions of the signal you are connecting to the 'mo' port of the M...

4 months ago | 0

| accepted