Publications

You can also find my articles on my Google Scholar profile.

Journal Articles

Multi-population differential evolution approach for feature selection with mutual information ranking

Published in Expert Systems with Applications, 2025

This paper proposes a novel multi-population differential evolution approach for feature selection with mutual information ranking, which significantly enhances classification performance by reducing feature dimensionality and improving algorithm optimization capabilities.

Recommended citation: Fei Yu, Jian Guan, Hongrun Wu, Hui Wang, Biyang Ma. (2024). "Multi-population differential evolution approach for feature selection with mutual information ranking." Expert Systems with Applications, 260, 125404.
Download Paper | Download Slides

Multi-Agent Transfer Reinforcement Learning for Resource Management in Underwater Acoustic Communication Networks

Published in IEEE Transactions on Network Science and Engineering, 2023

This paper presents a multi-agent transfer reinforcement learning approach for efficient resource management in underwater acoustic communication networks, aiming to improve network performance and adaptability.

Recommended citation: Wang, Hui; Wu, Hongrun; Chen, Yingpin; Ma, Biyang. (2023). "Multi-Agent Transfer Reinforcement Learning for Resource Management in Underwater Acoustic Communication Networks." IEEE Transactions on Network Science and Engineering.11(2), 2012-2023.
Download Paper | Download Slides

Tensor decomposition for multi-agent predictive state representation

Published in Expert Systems with Applications, 2022

This paper proposes tensor techniques to learn a multi-agent PSR model, addressing the challenges of increasing agent numbers and problem complexity, and demonstrating effectiveness across multiple domains.

Recommended citation: Biyang Ma, Bilian Chen, Yifeng Zeng*, Jing Tang, Langcai Cao (2022). " Tensor decomposition for multi-agent predictive state representation." Expert Systems with Applications, 2022,115969, 2022;. 189(1).
Download Paper | Download Slides

Tensor optimization with group lasso for multi-agent predictive state representation

Published in Knowledge-Based Systems, 2021

This paper proposes a tensor optimization approach to learn a multi-agent PSR model, addressing the challenges of limited samples and increasing number of agents, and demonstrating promising performance across multiple problem domains.

Recommended citation: Biyang Ma, Jing Tang, Bilian Chen, Yinghui Pan, Yifeng Zeng (2021). " Tensor optimization with group lasso for multi-agent predictive state representation." Knowledge-Based Systems, 2021, 106893., 2021. 211(7).
Download Paper | Download Slides

Conference Papers

Improvement and Evaluation of the Policy Legibility in Reinforcement Learning

Published in AAMAS, 2023

In this article, we propose a novel reward shaping mechanism to enhance the legibility of reinforcement learning policies for intelligent agents, and develop an interactive system to gather user evaluations and demonstrate the approach’s performance.

Recommended citation: Yanyu Liu, Yifeng Zeng, Biyang Ma, Yinghui Pan, Huifan Gao, Xiaohan Huang (2023). "Improvement and Evaluation of the Policy Legibility in Reinforcement Learning." In Proc. Of AAMAS ’23. 2023, 3044–3046.
Download Paper | Download Slides

LBfT: Learning Bayesian Network Structures from Text in Autonomous Typhoon Response Systems

Published in AAMAS, 2022

We demonstrate a deep learning framework that identifies typhoon-relevant variables and builds their causal relations from text, enhancing decision models in autonomous typhoon response systems using the CausalBank dataset and user domain knowledge.

Recommended citation: Yinghui Pan, Junhan Chen, Yifeng Zeng, Zhangrui Yao, Qianwen Li, Biyang Ma, Yi Ji, and Zhong Ming. (2022). "LBfT: Learning Bayesian Network Structures from Text in Autonomous Typhoon Response Systems." In Proc. Of AAMAS ’22. 2022, 1914–1916.
Download Paper | Download Slides

Ev-IDID: Enhancing solutions to interactive dynamic influence diagrams through evolutionary algorithms

Published in AAMAS, 2022

This demo introduces an interactive I-DID system that incorporates state-of-the-art and novel evolutionary algorithms, enabling users to specify parameters, visualize solutions, and automate behavioral model generation for multiagent sequential decision-making under uncertainty.

Recommended citation: Biyang Ma, Yinghui Pan, Yifeng Zeng, Zhong Ming.(2022). "Ev-IDID: Enhancing solutions to interactive dynamic influence diagrams through evolutionary algorithms " In Proc. Of AAMAS ’22. 2022, 1911–1913.
Download Paper | Download Slides

ATPT: Automate Typhoon Contingency Plan Generation from Text.

Published in AAMAS, 2021

We present and implement a framework that utilizes deep learning techniques to automate the generation of a planning domain model from natural language input, demonstrating its application in automatically generating typhoon contingency plans from official documents.

Recommended citation: Yifeng Zeng, Zhangrui Yao, Yinghui Pan, Wanqing Chen, Junxin Zhou, Junhan Chen, Biyang Ma, and Zhong Ming. (2021). "ATPT: Automate Typhoon Contingency Plan Generation from Text." In Proc. Of AAMAS ’21. 2021, 1788–1790.
Download Paper | Download Slides