Difference between revisions of "Matthia Sabatelli"
GerdIsenberg (talk | contribs) |
GerdIsenberg (talk | contribs) |
||
| (One intermediate revision by the same user not shown) | |||
| Line 7: | Line 7: | ||
researching on [https://en.wikipedia.org/wiki/Transfer_learning transfer learning] which bridges between [[Supervised Learning|supervised]] and [[Deep Learning|deep]] [[Reinforcement Learning|reinforcement learning]]. | researching on [https://en.wikipedia.org/wiki/Transfer_learning transfer learning] which bridges between [[Supervised Learning|supervised]] and [[Deep Learning|deep]] [[Reinforcement Learning|reinforcement learning]]. | ||
He holds a B.Sc. from [https://en.wikipedia.org/wiki/University_of_Trento University of Trento] in 2014, and a M.Sc. from [https://en.wikipedia.org/wiki/University_of_Groningen University of Groningen] in 2017 <ref>[https://www.linkedin.com/in/matthia-sabatelli-70370b93/ Matthia Sabatelli | LinkedIn]</ref>. | He holds a B.Sc. from [https://en.wikipedia.org/wiki/University_of_Trento University of Trento] in 2014, and a M.Sc. from [https://en.wikipedia.org/wiki/University_of_Groningen University of Groningen] in 2017 <ref>[https://www.linkedin.com/in/matthia-sabatelli-70370b93/ Matthia Sabatelli | LinkedIn]</ref>. | ||
| − | In his M.Sc thesis Matthia Sabatelli elaborates on [[Learning|learning]] to play chess with minimal [[Search|lookahead]], using [[Neural Networks#Deep|multilayer perceptrons]] versus [[Neural Networks#Convolutional|convolutional neural networks]] to approximate [[Stockfish|Stockfish’s]] [[Evaluation|evaluation]], also comparing two different [[Board Representation|board representations]] for the input layer | + | |
| + | =Chess= | ||
| + | At University of Groningen, Matthia Sabatelli worked on chess - the project work along with [[Zacharias Georgiou]], [[Evangelos Karountzos]] and [[Yaroslav Shkarupa]] dealt with [[Reinforcement Learning|reinforcement learning]] in simple [[Endgame|chess endgames]] such as [[KRK]] <ref>[[Zacharias Georgiou]], [[Evangelos Karountzos]], [[Yaroslav Shkarupa]], [[Matthia Sabatelli]] ('''2016'''). ''A Reinforcement Learning Approach for Solving KRK Chess Endgames''. [https://github.com/paintception/A-Reinforcement-Learning-Approach-for-Solving-Chess-Endgames/blob/master/project_papers/final_paper/reinforcement-learning-approach(2).pdf pdf]</ref>. | ||
| + | In his M.Sc thesis, supervised by [[Marco Wiering]] and [[Valeriu Codreanu]], Matthia Sabatelli elaborates on [[Learning|learning]] to play chess with minimal [[Search|lookahead]], using [[Neural Networks#Deep|multilayer perceptrons]] versus [[Neural Networks#Convolutional|convolutional neural networks]] to approximate [[Stockfish|Stockfish’s]] [[Evaluation|evaluation]], also comparing two different [[Board Representation|board representations]] for the input layer | ||
<ref>[[Matthia Sabatelli]] ('''2017'''). ''Learning to Play Chess with Minimal Lookahead and Deep Value Neural Networks''. Master's thesis, [https://en.wikipedia.org/wiki/University_of_Groningen University of Groningen], [https://www.ai.rug.nl/~mwiering/Thesis_Matthia_Sabatelli.pdf pdf]</ref>. | <ref>[[Matthia Sabatelli]] ('''2017'''). ''Learning to Play Chess with Minimal Lookahead and Deep Value Neural Networks''. Master's thesis, [https://en.wikipedia.org/wiki/University_of_Groningen University of Groningen], [https://www.ai.rug.nl/~mwiering/Thesis_Matthia_Sabatelli.pdf pdf]</ref>. | ||
| Line 13: | Line 16: | ||
<ref>[https://dblp.org/pid/160/6434.html dblp: Matthia Sabatelli]</ref> | <ref>[https://dblp.org/pid/160/6434.html dblp: Matthia Sabatelli]</ref> | ||
==2016 ...== | ==2016 ...== | ||
| − | * Zacharias Georgiou, Evangelos Karountzos, Yaroslav Shkarupa, [[Matthia Sabatelli]] ('''2016'''). ''A Reinforcement Learning Approach for Solving KRK Chess Endgames''. [https://github.com/paintception/A-Reinforcement-Learning-Approach-for-Solving-Chess-Endgames/blob/master/project_papers/final_paper/reinforcement-learning-approach(2).pdf pdf] <ref>[https://github.com/paintception/A-Reinforcement-Learning-Approach-for-Solving-Chess-Endgames GitHub - paintception/A-Reinforcement-Learning-Approach-for-Solving-Chess-Endgames: Machine Learning - Reinforcement Learning]</ref> | + | * [[Zacharias Georgiou]], [[Evangelos Karountzos]], [[Yaroslav Shkarupa]], [[Matthia Sabatelli]] ('''2016'''). ''A Reinforcement Learning Approach for Solving KRK Chess Endgames''. [https://github.com/paintception/A-Reinforcement-Learning-Approach-for-Solving-Chess-Endgames/blob/master/project_papers/final_paper/reinforcement-learning-approach(2).pdf pdf] <ref>[https://github.com/paintception/A-Reinforcement-Learning-Approach-for-Solving-Chess-Endgames GitHub - paintception/A-Reinforcement-Learning-Approach-for-Solving-Chess-Endgames: Machine Learning - Reinforcement Learning]</ref> |
* [[Matthia Sabatelli]] ('''2017'''). ''Learning to Play Chess with Minimal Lookahead and Deep Value Neural Networks''. Master's thesis, [https://en.wikipedia.org/wiki/University_of_Groningen University of Groningen], [https://www.ai.rug.nl/~mwiering/Thesis_Matthia_Sabatelli.pdf pdf] <ref>[https://github.com/paintception/DeepChess GitHub - paintception/DeepChess]</ref> | * [[Matthia Sabatelli]] ('''2017'''). ''Learning to Play Chess with Minimal Lookahead and Deep Value Neural Networks''. Master's thesis, [https://en.wikipedia.org/wiki/University_of_Groningen University of Groningen], [https://www.ai.rug.nl/~mwiering/Thesis_Matthia_Sabatelli.pdf pdf] <ref>[https://github.com/paintception/DeepChess GitHub - paintception/DeepChess]</ref> | ||
* [[Matthia Sabatelli]], [[Francesco Bidoia]], [[Valeriu Codreanu]], [[Marco Wiering]] ('''2018'''). ''Learning to Evaluate Chess Positions with Deep Neural Networks and Limited Lookahead''. ICPRAM 2018, [https://www.ai.rug.nl/~mwiering/GROUP/ARTICLES/ICPRAM_CHESS_DNN_2018.pdf pdf] | * [[Matthia Sabatelli]], [[Francesco Bidoia]], [[Valeriu Codreanu]], [[Marco Wiering]] ('''2018'''). ''Learning to Evaluate Chess Positions with Deep Neural Networks and Limited Lookahead''. ICPRAM 2018, [https://www.ai.rug.nl/~mwiering/GROUP/ARTICLES/ICPRAM_CHESS_DNN_2018.pdf pdf] | ||
Latest revision as of 21:53, 27 May 2021
Home * People * Matthia Sabatelli
Matthia Sabatelli [1]
Matthia Sabatelli,
an Italian computer scientist and Ph.D. candidate at University of Liège,
researching on transfer learning which bridges between supervised and deep reinforcement learning.
He holds a B.Sc. from University of Trento in 2014, and a M.Sc. from University of Groningen in 2017 [2].
Chess
At University of Groningen, Matthia Sabatelli worked on chess - the project work along with Zacharias Georgiou, Evangelos Karountzos and Yaroslav Shkarupa dealt with reinforcement learning in simple chess endgames such as KRK [3]. In his M.Sc thesis, supervised by Marco Wiering and Valeriu Codreanu, Matthia Sabatelli elaborates on learning to play chess with minimal lookahead, using multilayer perceptrons versus convolutional neural networks to approximate Stockfish’s evaluation, also comparing two different board representations for the input layer [4].
Selected Publications
2016 ...
- Zacharias Georgiou, Evangelos Karountzos, Yaroslav Shkarupa, Matthia Sabatelli (2016). A Reinforcement Learning Approach for Solving KRK Chess Endgames. pdf [6]
- Matthia Sabatelli (2017). Learning to Play Chess with Minimal Lookahead and Deep Value Neural Networks. Master's thesis, University of Groningen, pdf [7]
- Matthia Sabatelli, Francesco Bidoia, Valeriu Codreanu, Marco Wiering (2018). Learning to Evaluate Chess Positions with Deep Neural Networks and Limited Lookahead. ICPRAM 2018, pdf
- Matthia Sabatelli, Gilles Louppe, Pierre Geurts, Marco Wiering (2018). Deep Quality-Value (DQV) Learning. arXiv:1810.00368
- Matthia Sabatelli, Gilles Louppe, Pierre Geurts, Marco Wiering (2019). Approximating two value functions instead of one: towards characterizing a new family of Deep Reinforcement Learning algorithms. arXiv:1909.01779
2020 ...
- Matthia Sabatelli, Gilles Louppe, Pierre Geurts, Marco Wiering (2020). The Deep Quality-Value Family of Deep Reinforcement Learning Algorithms. IJCNN 2020 [8]
- Matthia Sabatelli, Mike Kestemont, Pierre Geurts (2020). On the Transferability of Winning Tickets in Non-Natural Image Datasets. arXiv:2005.05232
External Links
- Matthia Sabatelli | LinkedIn
- Matthia Sabatelli
- Matthia Sabatelli - Google Scholar
- paintception (Matthia Sabatelli) · GitHub
References
- ↑ Matthia Sabatelli
- ↑ Matthia Sabatelli | LinkedIn
- ↑ Zacharias Georgiou, Evangelos Karountzos, Yaroslav Shkarupa, Matthia Sabatelli (2016). A Reinforcement Learning Approach for Solving KRK Chess Endgames. pdf
- ↑ Matthia Sabatelli (2017). Learning to Play Chess with Minimal Lookahead and Deep Value Neural Networks. Master's thesis, University of Groningen, pdf
- ↑ dblp: Matthia Sabatelli
- ↑ GitHub - paintception/A-Reinforcement-Learning-Approach-for-Solving-Chess-Endgames: Machine Learning - Reinforcement Learning
- ↑ GitHub - paintception/DeepChess
- ↑ GitHub - paintception/Deep-Quality-Value-DQV-Learning-: DQV-Learning: a novel faster synchronous Deep Reinforcement Learning algorithm
