Restarting
Yes, I have resumed the TFM. Since so much time has passed, I have been forced to review all the technology I was using. Taking advantage of the situation, I...
Posts by Tag
log
Week 8-9. Q learning - fixes
Since the last post I have focused more on solving some bugs and preparing the execution environment of the algorithm in cloud.
Week 6-7. Q learning - new hardware
For this week we are still trying to find maximum speeds with the car running the software on my local computer. But this week I managed to get access to the...
Week 6-7. Q learning - Playing with parameters
Observing the large training time that the algorithm needs to learn to solve the problem when the iteration speed is limited (since the lower the speed, the ...
Week 6-7. Q learning - introducing stop and go and iteration velocity
On this occasion, it has been decided to improve the results obtained on the previous occasion. For this purpose, two techniques have been used to squeeze th...
Week 6-7. Q learning - completed
The problem has been solved with the Q Learning algorithm parameterized with discrete action space of size 16 (4 bins for linear velocity and 4 for angular v...
Week 6-7. Q learning - changes
I am still working on how to improve the handling when the problem is multivariable (linear and angular velocity) by applying the Q Learning algorithm. The c...
Week 5-6. Simple Q learning - completed
For today I wanted to show a video with a small demonstration of how the implemented algorithm has managed to solve the problem. It should be noted that the ...
Week 5-6. Simple Q learning - new reward function and circuit completed
For today I managed to get the car to drive around the circuit at a constant speed. This has required some changes to the reward function I had so far. The c...
Week 5-6. Simple Q learning - updated
In the last post I published an implementation of the simplified Q learning algorithm for this problem in which I discretised both the space of observations ...
Week 5-6. Simple Q learning
As a first approach to the problem at hand, I proceed to build a new environment from the original MontmeloLine environment. In this new environment the acti...
Week 5-6. Building Gym Gazebo environments
In the week, a Formula 1 training environment based on Gazebo and ROS Noetic has been introduced. These environments allow agent training based on reinforcem...
Week 1-4. Learning about reinforcement learning
This month I was studying reinforcement learning and his implementation in Python. I was using Open AI Gym (a framework for setting the environement where th...
Week 0. Start the trip!
I started reading some books adbout Machine Learning. I prefer to focus on books about reinforcement learning and deep reinforcement learnig because I think ...
update
Restarting
Yes, I have resumed the TFM. Since so much time has passed, I have been forced to review all the technology I was using. Taking advantage of the situation, I...
Week 8-9. Q learning - fixes
Since the last post I have focused more on solving some bugs and preparing the execution environment of the algorithm in cloud.
Week 6-7. Q learning - new hardware
For this week we are still trying to find maximum speeds with the car running the software on my local computer. But this week I managed to get access to the...
Week 6-7. Q learning - Playing with parameters
Observing the large training time that the algorithm needs to learn to solve the problem when the iteration speed is limited (since the lower the speed, the ...
Week 6-7. Q learning - introducing stop and go and iteration velocity
On this occasion, it has been decided to improve the results obtained on the previous occasion. For this purpose, two techniques have been used to squeeze th...
Week 6-7. Q learning - completed
The problem has been solved with the Q Learning algorithm parameterized with discrete action space of size 16 (4 bins for linear velocity and 4 for angular v...
Week 6-7. Q learning - changes
I am still working on how to improve the handling when the problem is multivariable (linear and angular velocity) by applying the Q Learning algorithm. The c...
Week 5-6. Simple Q learning - completed
For today I wanted to show a video with a small demonstration of how the implemented algorithm has managed to solve the problem. It should be noted that the ...
Week 5-6. Simple Q learning - new reward function and circuit completed
For today I managed to get the car to drive around the circuit at a constant speed. This has required some changes to the reward function I had so far. The c...
Week 5-6. Simple Q learning - updated
In the last post I published an implementation of the simplified Q learning algorithm for this problem in which I discretised both the space of observations ...
Week 5-6. Simple Q learning
As a first approach to the problem at hand, I proceed to build a new environment from the original MontmeloLine environment. In this new environment the acti...
Week 5-6. Building Gym Gazebo environments
In the week, a Formula 1 training environment based on Gazebo and ROS Noetic has been introduced. These environments allow agent training based on reinforcem...
Week 1-4. Learning about reinforcement learning
This month I was studying reinforcement learning and his implementation in Python. I was using Open AI Gym (a framework for setting the environement where th...
Week 0. Start the trip!
I started reading some books adbout Machine Learning. I prefer to focus on books about reinforcement learning and deep reinforcement learnig because I think ...