brozek/website
1
0
Fork 0
mirror of https://github.com/Brandon-Rozek/website.git synced 2025-10-09 06:21:13 +00:00
Code Issues Projects Releases Packages Activity

Updated research page

Browse source
This commit is contained in:
Brandon Rozek 2022-02-17 13:37:46 -05:00
parent 93e42953e0
commit a12db74b98
3 changed files with 46 additions and 22 deletions
Download patch file Download diff file Expand all files Collapse all files

16
content/research/deepreinforcementlearning/_index.md
View file

@ -3,20 +3,22 @@ Title: Deep Reinforcement Learning
Description: Combining Reinforcement Learning with Deep Learning
---
In the Fall of 2019, I look at integrating demonstration data into a reinforcement learning algorithm in order to make it sample efficient.
I am interested in sample-efficient reinforcement learning.
That is, decreases the number of interactions an agent needs
with an environment to achieve some goal. In the Fall of 2019,
I approached this by integrating interactive demonstration
data into the optimized Deep Q-Networks algorithm.
The results are positive and are heavily documented through the following:
[Honors Thesis](/files/research/honorsthesis.pdf)
[Undergraduate Honors Thesis](/files/research/honorsthesis.pdf)
[Honors Defense](/files/research/ExpeditedLearningInteractiveDemo.pptx)
[Undergraduate Honors Defense](/files/research/ExpeditedLearningInteractiveDemo.pptx)
Thanks to my advisor Dr. Ron Zacharksi and my committee members for all their feedback on my work!
In the spring of 2019, under the guidance of Dr. Ron Zacharski I practiced several of the modern techniques used in Reinforcement Learning today.
The semester prior, I built a [reinforcement learning library](https://github.com/brandon-rozek/rltorch) with implementations of several popular papers. ([Semi-Weekly Progress](weeklyprogress)).
I facilitated my learning by creating a [reinforcement learning library](https://github.com/brandon-rozek/rltorch) with implementations of several popular papers. ([Semi-Weekly Progress](weeklyprogress))
I also presented my research (which involved creating an algorithm) at my school's research symposium. ([Slides](/files/research/QEP.pptx)) ([Abstract](abstractspring2019))
I also presented at my school's research symposium. ([Slides](/files/research/QEP.pptx)) ([Abstract](abstractspring2019))
In the summer of 2019, I became interested in having the interactions with the environment be in a separate process. This inspired two different implementations, [ZeroMQ](https://github.com/brandon-rozek/zerogym) and [HTTP](https://github.com/brandon-rozek/gymhttp). Given the option, you should use the ZeroMQ implementation since it contains less communication overhead.