Deep learning and computer chess (Part 1): using neural networks for chess evaluation functions
This report presents the implementation of two different chess evaluation functions based on the Giraffe and DeepChess papers. In the first implementation, the evaluator network architecture from Giraffe’s evaluation function was adapted into a multiclass classifier designed to predict 7 classi...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project (FYP) |
| Language: | English |
| Published: |
Nanyang Technological University
2024
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/175276 |
| _version_ | 1811683358317477888 |
|---|---|
| author | U, Jeremy Keat |
| author2 | He Ying |
| author_facet | He Ying U, Jeremy Keat |
| author_sort | U, Jeremy Keat |
| collection | NTU |
| description | This report presents the implementation of two different chess evaluation functions based on
the Giraffe and DeepChess papers. In the first implementation, the evaluator network
architecture from Giraffe’s evaluation function was adapted into a multiclass classifier
designed to predict 7 classifications of Stockfish evaluations through supervised learning.
Experiments were conducted to gauge the effectiveness of input feature representations and
dropout regularisation. The second implementation, based on DeepChess, uses a different
approach to evaluation, through comparison of two chess positions in a Siamese network and
outputs which of the two has a more advantageous position, evaluating board positions
through binary classification. The network was trained in a two-stage process with a
combination of unsupervised and supervised learning. Experiments were conducted to
observe the effect of freezing pretrained layer weights as well as changing layer activation
functions to LeakyReLU. |
| first_indexed | 2024-10-01T04:11:28Z |
| format | Final Year Project (FYP) |
| id | ntu-10356/175276 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T04:11:28Z |
| publishDate | 2024 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1752762024-04-26T15:44:17Z Deep learning and computer chess (Part 1): using neural networks for chess evaluation functions U, Jeremy Keat He Ying School of Computer Science and Engineering YHe@ntu.edu.sg Computer and Information Science This report presents the implementation of two different chess evaluation functions based on the Giraffe and DeepChess papers. In the first implementation, the evaluator network architecture from Giraffe’s evaluation function was adapted into a multiclass classifier designed to predict 7 classifications of Stockfish evaluations through supervised learning. Experiments were conducted to gauge the effectiveness of input feature representations and dropout regularisation. The second implementation, based on DeepChess, uses a different approach to evaluation, through comparison of two chess positions in a Siamese network and outputs which of the two has a more advantageous position, evaluating board positions through binary classification. The network was trained in a two-stage process with a combination of unsupervised and supervised learning. Experiments were conducted to observe the effect of freezing pretrained layer weights as well as changing layer activation functions to LeakyReLU. Bachelor's degree 2024-04-23T05:46:37Z 2024-04-23T05:46:37Z 2024 Final Year Project (FYP) U, J. K. (2024). Deep learning and computer chess (Part 1): using neural networks for chess evaluation functions. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/175276 https://hdl.handle.net/10356/175276 en application/pdf Nanyang Technological University |
| spellingShingle | Computer and Information Science U, Jeremy Keat Deep learning and computer chess (Part 1): using neural networks for chess evaluation functions |
| title | Deep learning and computer chess (Part 1): using neural networks for chess evaluation functions |
| title_full | Deep learning and computer chess (Part 1): using neural networks for chess evaluation functions |
| title_fullStr | Deep learning and computer chess (Part 1): using neural networks for chess evaluation functions |
| title_full_unstemmed | Deep learning and computer chess (Part 1): using neural networks for chess evaluation functions |
| title_short | Deep learning and computer chess (Part 1): using neural networks for chess evaluation functions |
| title_sort | deep learning and computer chess part 1 using neural networks for chess evaluation functions |
| topic | Computer and Information Science |
| url | https://hdl.handle.net/10356/175276 |
| work_keys_str_mv | AT ujeremykeat deeplearningandcomputerchesspart1usingneuralnetworksforchessevaluationfunctions |