Chi-Jen Lu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Chi-Jen Lu has a total of 29 co-authors. Here are the top co-authors by number of shared publications:
- Chuang-Chieh Lin
National Taiwan Ocean University, Zhongzheng Dist, Keelung City, Taiwan
6 shared publications - Po-An Chen
National yang Ming Chiao Tung University
5 shared publications - Shi-Chun Tsai
National Chiao Tung University
5 shared publications - Po-An Chen
Institute of Information Management National Yang-Ming Chiao-Tung University 1001 University Rd 300 Hsinchu City Taiwan
4 shared publications - Chih-Chieh Hung
Department of Management Information Systems, National Chung Hsing University, Taichung City, Taiwan
3 shared publications Chen-Yu Wei
Institute of Information Science Academia Sinica Taiwan
2 shared publications
Latest Publications
- Open Access
- 24 February 2026
- Published by Elsevier in SSRN Electronic Journal
Strategic Nominations in Election Games: Equilibria, Complexity, and Inefficiency
We study election games in which each group nominates one candidate and receives as payoff the expected utility of its members under a probabilistic winning rule. Motivated by simulations of two- and three-group elections, we adopt a monotonicity assumption relating a candidate’s social utility to its winning probability. For three or more groups, we show that pure-strategy Nash equilibria (PSNE) may fail to exist even under egoistic preferences, and that deciding PSNE existence is NP-complete in a succinct (general-form) representation. Under a cross-monotone winning-probability function, we give simple sufficient conditions for PSNE existence and an FPT algorithm to compute a PSNE, parameterized by the number of irresolute groups and the nominating depth. Finally, under cross-monotonicity and an order-preserving condition, we bound the price of anarchy of egoistic election games by the number of groups, and show that this bound is tight for the hardmax winning-probability rule.
- 24 February 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessStrategic Nominations in Election Games: Equilibria, Complexity, and Inefficiency
We study election games in which each group nominates one candidate and receives as payoff the expected utility of its members under a probabilistic winning rule. Motivated by simulations of two- and three-group elections, we adopt a monotonicity assumption relating a candidate’s social utility to its winning probability. For three or more groups, we show that pure-strategy Nash equilibria (PSNE) may fail to exist even under egoistic preferences, and that deciding PSNE existence is NP-complete in a succinct (general-form) representation. Under a cross-monotone winning-probability function, we give simple sufficient conditions for PSNE existence and an FPT algorithm to compute a PSNE, parameterized by the number of irresolute groups and the nominating depth. Finally, under cross-monotonicity and an order-preserving condition, we bound the price of anarchy of egoistic election games by the number of groups, and show that this bound is tight for the hardmax winning-probability rule.
- Open Access
- 29 October 2025
- Published by Springer Nature in Research Square
A Computational Analysis of Strategic Nominations: Modeling Equilibrium and Complexity in Organizational Elections
This paper models organizational electoral competition as an election game with strategic organizations in a macro level. Members of the organization derive utility from the elected candidate. Each organization nominates exactly one candidate to compete against candidates from other organizations and the payoff of each organization is the expected utility its members receive. By simulating the voting outcomes in the competitions between two and three organizations, we reveal and validate the monotone correspondence between the utility received by the voters and the chance of winning in the competition. Next, we show that pure-strategy Nash equilibria (PSNE) do not always exist for the election games with three or more groups, even under egoistic conditions, and prove that determining the PSNE existence is {\sf NP}-complete in general-form representation. Addressing this, we propose sufficient conditions for PSNE existence and develop a fixed-parameter tractable algorithm to compute equilibria, parameterized by irresolute groups and nominating depth. We also show that the price of anarchy for egoistic election games is upper bounded by the number of competing groups. These findings highlight how stability and efficiency can deteriorate with increased group participation, offering computational insights into multi-group strategic decision-making.
- 29 October 2025
- Published by Springer Nature in Research Square
Open AccessA Computational Analysis of Strategic Nominations: Modeling Equilibrium and Complexity in Organizational Elections
This paper models organizational electoral competition as an election game with strategic organizations in a macro level. Members of the organization derive utility from the elected candidate. Each organization nominates exactly one candidate to compete against candidates from other organizations and the payoff of each organization is the expected utility its members receive. By simulating the voting outcomes in the competitions between two and three organizations, we reveal and validate the monotone correspondence between the utility received by the voters and the chance of winning in the competition. Next, we show that pure-strategy Nash equilibria (PSNE) do not always exist for the election games with three or more groups, even under egoistic conditions, and prove that determining the PSNE existence is {\sf NP}-complete in general-form representation. Addressing this, we propose sufficient conditions for PSNE existence and develop a fixed-parameter tractable algorithm to compute equilibria, parameterized by irresolute groups and nominating depth. We also show that the price of anarchy for egoistic election games is upper bounded by the number of competing groups. These findings highlight how stability and efficiency can deteriorate with increased group participation, offering computational insights into multi-group strategic decision-making.
- 19 August 2024
- Published by Elsevier in Theoretical Computer Science
- Vol. 1017
Multiagent learning for competitive opinion optimization
- 19 August 2024
- Published by Elsevier in Theoretical Computer Science
- Vol. 1017
Multiagent learning for competitive opinion optimization
- 17 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computational Intelligence Magazine
- Vol. 18 (4) , 60-61
Group Formation by Group Joining and Opinion Updates Via Multi-Agent Online Gradient Ascent [AI-eXplained]
This article aims to exemplify best-response dynamics and multi-agent online learning by group formation. This extended abstract provides a summary of the full paper in IEEE Computational Intelligence Magazine on the special issue AI-eXplained (AI-X). The full paper includes interactive components to facilitate interested readers to grasp the idea of pure-strategy Nash equilibria and how the system of strategic agents converges to a stable state by the decentralized online gradient ascent with and without regularization.
- 17 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computational Intelligence Magazine
- Vol. 18 (4) , 60-61
Group Formation by Group Joining and Opinion Updates Via Multi-Agent Online Gradient Ascent [AI-eXplained]
This article aims to exemplify best-response dynamics and multi-agent online learning by group formation. This extended abstract provides a summary of the full paper in IEEE Computational Intelligence Magazine on the special issue AI-eXplained (AI-X). The full paper includes interactive components to facilitate interested readers to grasp the idea of pure-strategy Nash equilibria and how the system of strategic agents converges to a stable state by the decentralized online gradient ascent with and without regularization.
- 3 June 2023
- Published by Springer Nature
- p. 392-403
Budget-Constrained Cost-Covering Job Assignment for a Total Contribution-Maximizing Platform
- 3 June 2023
- Published by Springer Nature
- p. 392-403
Budget-Constrained Cost-Covering Job Assignment for a Total Contribution-Maximizing Platform
- Open Access
Multiagent Learning for Competitive Opinion Optimization
From a perspective of designing or engineering for opinion formation games in social networks, the "opinion maximization (or minimization)" problem has been studied mainly for designing subset selecting algorithms. We define a two-player zero-sum Stackelberg game of competitive opinion optimization by letting the player under study as the leader minimize the sum of expressed opinions by doing so-called "internal opinion design", knowing that the other adversarial player as the follower is to maximize the same objective by also conducting her own internal opinion design. We furthermore consider multiagent learning, specifically using the Optimistic Gradient Descent Ascent, and analyze its convergence to equilibria in the simultaneous version of competitive opinion optimization.
Open AccessMultiagent Learning for Competitive Opinion Optimization
From a perspective of designing or engineering for opinion formation games in social networks, the "opinion maximization (or minimization)" problem has been studied mainly for designing subset selecting algorithms. We define a two-player zero-sum Stackelberg game of competitive opinion optimization by letting the player under study as the leader minimize the sum of expressed opinions by doing so-called "internal opinion design", knowing that the other adversarial player as the follower is to maximize the same objective by also conducting her own internal opinion design. We furthermore consider multiagent learning, specifically using the Optimistic Gradient Descent Ascent, and analyze its convergence to equilibria in the simultaneous version of competitive opinion optimization.
- 1 January 2022
- Published by Springer Nature
- p. 61-72
Multiagent Learning for Competitive Opinion Optimization (Extended Abstract)
- 1 January 2022
- Published by Springer Nature
- p. 61-72
Multiagent Learning for Competitive Opinion Optimization (Extended Abstract)
- Open Access
Rethinking Why Intermediate-Task Fine-Tuning Works
Supplementary Training on Intermediate Labeled-data Tasks (STILTs) is a widely applied technique, which first fine-tunes the pretrained language models on an intermediate task before on the target task of interest. While STILTs is able to further improve the performance of pretrained language models, it is still unclear why and when it works. Previous research shows that those intermediate tasks involving complex inference, such as commonsense reasoning, work especially well for RoBERTa. In this paper, we discover that the improvement from an intermediate task could be orthogonal to it containing reasoning or other complex skills -- a simple real-fake discrimination task synthesized by GPT2 can benefit diverse target tasks. We conduct extensive experiments to study the impact of different factors on STILTs. These findings suggest rethinking the role of intermediate fine-tuning in the STILTs pipeline.
Open AccessRethinking Why Intermediate-Task Fine-Tuning Works
Supplementary Training on Intermediate Labeled-data Tasks (STILTs) is a widely applied technique, which first fine-tunes the pretrained language models on an intermediate task before on the target task of interest. While STILTs is able to further improve the performance of pretrained language models, it is still unclear why and when it works. Previous research shows that those intermediate tasks involving complex inference, such as commonsense reasoning, work especially well for RoBERTa. In this paper, we discover that the improvement from an intermediate task could be orthogonal to it containing reasoning or other complex skills -- a simple real-fake discrimination task synthesized by GPT2 can benefit diverse target tasks. We conduct extensive experiments to study the impact of different factors on STILTs. These findings suggest rethinking the role of intermediate fine-tuning in the STILTs pipeline.
- 7 July 2021
- Published by Springer Nature in Theory of Computing Systems
- Vol. 66 (1) , 38-55
An Alternating Algorithm for Finding Linear Arrow-Debreu Market Equilibria
- 7 July 2021
- Published by Springer Nature in Theory of Computing Systems
- Vol. 66 (1) , 38-55
An Alternating Algorithm for Finding Linear Arrow-Debreu Market Equilibria
- 25 February 2021
- Published by Springer Nature
- p. 509-525
TinyGAN: Distilling BigGAN for Conditional Image Generation
- 25 February 2021
- Published by Springer Nature
- p. 509-525
TinyGAN: Distilling BigGAN for Conditional Image Generation