https://github.com/kirillbobyrev/td-gamma

Empirical Study of TD(γ) Reinforcement Learning algorithm for Value function Estimation
https://github.com/kirillbobyrev/td-gamma

class-project optimization poster reinforcement-learning

Last synced: 3 months ago
JSON representation

Empirical Study of TD(γ) Reinforcement Learning algorithm for Value function Estimation

• Host: GitHub
• URL: https://github.com/kirillbobyrev/td-gamma
• Owner: kirillbobyrev
• License: apache-2.0
• Created: 2018-03-13T18:47:18.000Z (over 8 years ago)
• Default Branch: main
• Last Pushed: 2021-04-25T14:21:58.000Z (about 5 years ago)
• Last Synced: 2025-01-01T03:14:07.835Z (over 1 year ago)
• Topics: class-project, optimization, poster, reinforcement-learning
• Language: TeX
• Homepage: https://kbobyrev.github.io/resources/TD-Gamma-Poster.pdf
• Size: 2.38 MB
• Stars: 1
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# TD(γ)

Empirical Study of **TD(γ)** Reinforcement Learning algorithm for Value

function Estimation.

The studied algorithm was introduced in [TDγ: Re-evaluating Complex Backups in

Temporal Difference

Learning](https://papers.nips.cc/paper/4472-td_gamma-re-evaluating-complex-backups-in-temporal-difference-learning.pdf)

paper by G. Konidaris, S. Niekum and P. Thomas which was presented at NIPS

2011.

Obtained results show that **TD(γ)** was able to outperform its precursors

given a fairly simple environment and policy, but was much more

resource-intensive in terms of memory and computational complexity.

* [Compiled poster](https://kirillbobyrev.github.io/resources/TD-Gamma-Poster.pdf)

* [Code](https://colab.research.google.com/gist/kirillbobyrev/90efb074a4740f8a8a6abf22e58d988a/td-gamma.ipynb)