Jr-Chang Chen

Jr-Chang Chen

Auto-generated profile, based on publicly available information

Jr-Chang Chen

National Taipei University

Research Focus & Metrics

Academic Career Length

15Years (2012-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
5 February 2018 in IEEE Transactions on Games
Verified
Journal Article
Closed access
JCJr-Chang Chen
Jr-Chang Chen
GFGang-Yu Fan
Gang-Yu Fan
HCHung-Jui Chang
Hung-Jui Chang
THTsan-sheng Hsu
Tsan-sheng Hsu
Endgame databases can be difficult to use in tree search when database sizes remain large even after compression. Given the same endgame, we discover that the compression ratios vary significantly when using different encoding schemes. The intuition is that when a set of positions mapped into a continuous chunk of segments have similar values, block-based compression libraries such as zlib can yield a better compression ratio than cases where segments contain diversified values. However, finding the optimal encoding scheme by exhaustive enumeration is time-infeasible for endgame databases with a large number of pieces. We propose a novel approach using deep learning to obtain an encoding scheme so that the compression ratio is suitable for practical purposes. Our approach can be applied to chess-like games.

16 January 2019 in IEEE Transactions on Games
Verified
Journal Article
Closed access
JCJr-Chang Chen
Jr-Chang Chen
WTWen-Jie Tseng
Wen-Jie Tseng
IWI-Chen Wu
I-Chen Wu
TWTing-Han Wei
Ting-Han Wei
This paper describes the application of a modified comparison training for automatic feature weight tuning. The final objective is to improve the evaluation functions used in Chinese chess programs. First, we apply n-tuple networks to extract features. N-tuple networks require very little expert knowledge through its large numbers of features, while simultaneously allowing easy access. Second, we propose a modified comparison training into which tapered eval is incorporated. Experiments show that with the same features and the same Chinese chess program, the automatically tuned feature weights achieved a win rate of 86.58% against the hand-tuned ones. The above trained version was then improved by adding additional features, most importantly n-tuple features. This improved version achieved a win rate of 81.65% against the trained version without additional features.

29 July 2019 in IEEE Transactions on Games
Verified
Journal Article
Closed access
CLChing-Nung Lin
Ching-Nung Lin
JCJr-Chang Chen
Jr-Chang Chen
SYShi-jim Yen
Shi-jim Yen
Deep learning has become a hot topic in recent years. There are many teaching frameworks that ease the education process for deep learning. However, most current teaching examples either require a lot of training time or do not have interaction with users. Usually, the testing accuracy is the only evaluation criterion. However, it does not mean too much for a novice. We provide a framework which can teach the deep learning concepts with limited time or computational resource constraints. It offers students quick feedback, that is, the trained model can be easily tested on this framework. Students can increase training data, design a deep learning model or modify input planes to learn more advanced topics. The framework provides a Graphical User Interface, which connects to the trained model directly. It is a well-designed deep learning competition framework. This framework has been applied in the “Artificial Intelligence” undergraduate course for three semesters. The student feedback was quite positive. Furthermore, based on this framework, Deep Learning Othello competitions were held in TAAI and TCGA international computer game tournaments in 2017. We hope that this framework can encourage more people to be engaged in the research of deep learning and board games.

25 November 2023
Verified
Conference Paper
Open access
JCJr-Chang Chen
Jr-Chang Chen
THTzu-Yang Hsu
Tzu-Yang Hsu
CHChia-Ming Hsu
Chia-Ming Hsu
THTsan-sheng Hsu
Tsan-sheng Hsu
EinStein würfelt nicht! (abbr. EWN) is a stochastic game with a die rolled in each ply which makes it full of uncertainty. It does not look easy to design a good game engine using traditional alpha-beta-based approaches. In this paper, an EWN game engine is formed by combining n-tuple networks, trained by features extracted from two billion random games, with Monte Carlo Tree Search. Similar work has been proposed previously, but in this work, we compare performances on versions with n = 4, 6 and 9. We discover that 4-tuple networks add no extra strength in comparison with a plain baseline. However, a larger n does bring extra benefit. The jump of performance is noticeable from n = 4 to n = 6 and is very visible from n = 6 to n = 9. The tuple’s size n increases from 4 and 6 to 9. When n = 9, the size of the feature set is much larger than before and cannot be stored plainly in the main memory. One of our main contributions is to come out with a good encoding design taking care of various symmetry or equivalence properties so that it can be used in real-time game playing. Using our scheme, the number of distinct features is reduced to 20,244,124,980, which is 21.21% less than the original 95,440,494,357 features for 9-tuple networks.

21 February 2023 in IEEE Access
Verified
Journal Article
Open access
Chu-Hsuan HsuehChu-Hsuan Hsueh
Chu-Hsuan Hsueh
KIKokolo Ikeda
Kokolo Ikeda
IWI-Chen Wu
I-Chen Wu
JCJr-Chang Chen
Jr-Chang Chen
THTsan-Sheng Hsu
Tsan-Sheng Hsu
The AlphaZero algorithm achieved superhuman levels of play in chess, shogi, and Go, by learning without domain-specific knowledge except for game rules. This paper targets stochastic games and investigates whether AlphaZero can learn theoretical values and optimal play. Since the theoretical values of stochastic games are expected win rates, not a simple win, loss, or draw, it is worth investigating the ability of AlphaZero to approximate expected win rates of positions. This paper also thoroughly studies how AlphaZero is influenced by hyper-parameters and some implementation details. The analyses are mainly based on AlphaZero learning with lookup tables. Deep neural networks (DNNs) like the ones in the original AlphaZero are also experimented and compared. The tested stochastic games include reduced and strongly-solved variants of Chinese dark chess and EinStein würfelt nicht!. The experiments show that AlphaZero can learn policies that play almost optimally against the optimal player and can learn values accurately. In more detail, such good results are achieved by different hyper-parameter settings in a wide range, though it is observed that games on larger scales tend to have a little narrower range of proper hyper-parameters. In addition, the results of learning with DNNs are similar to lookup tables.

23 January 2018 in arXiv
Verified
Preprint
Open access
PDF file
WTWen-Jie Tseng
Wen-Jie Tseng
JCJr-Chang Chen
Jr-Chang Chen
IWI-Chen Wu
I-Chen Wu
TWTinghan Wei
Tinghan Wei
This paper describes the application of comparison training (CT) for automatic feature weight tuning, with the final objective of improving the evaluation functions used in Chinese chess programs. First, we propose an n-tuple network to extract features, since n-tuple networks require very little expert knowledge through its large numbers of features, while simulta-neously allowing easy access. Second, we propose a novel evalua-tion method that incorporates tapered eval into CT. Experiments show that with the same features and the same Chinese chess program, the automatically tuned comparison training feature weights achieved a win rate of 86.58% against the weights that were hand-tuned. The above trained version was then improved by adding additional features, most importantly n-tuple features. This improved version achieved a win rate of 81.65% against the trained version without additional features.