检索结果-内蒙古大学图书馆

To Analyze and Regulate Human-in-the-loop Learning for Congestion Games

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Li, Hongbo Duan, Lingjie The Pillar of Engineering Systems and Design Singapore University of Technology and Design 487372 Singapore

In congestion games, selfish users behave myopically to crowd to the shortest paths, and the social planner designs mechanisms to regulate such selfish routing through information or payment incentives. However, such mechanism design requires the knowledge of time-varying traffic conditions and it is the users themselves to learn and report past road experiences to the social planner (e.g., Waze or Google Maps). When congestion games meet mobile crowdsourcing, it is critical to incentivize selfish users to explore non-shortest paths in the best exploitation-exploration trade-off. First, we consider a simple but fundamental parallel routing network with one deterministic path and multiple stochastic paths for users with an average arrival probability λ. We prove that the current myopic routing policy (widely used in Waze and Google Maps) misses both exploration (when strong hazard belief) and exploitation (when weak hazard belief) as compared to the social optimum. Due to the myopic policy's under-exploration, we prove that the caused price of anarchy (PoA) is larger than (Equation presented) which can be arbitrarily large as discount factor ρ → 1. To mitigate such huge efficiency loss, we propose a novel selective information disclosure (SID) mechanism: we only reveal the latest traffic information to users when they intend to over-explore stochastic paths upon arrival, while hiding such information when they want to under-explore. We prove that our mechanism successfully reduces PoA to be less than 2. Besides the parallel routing network, we further extend our mechanism and PoA results to any linear path graphs with multiple intermediate nodes. In addition to the worst-case performance evaluation, we conduct extensive simulations with both synthetic and real transportation datasets to demonstrate the close-to-optimal average-case performance of our SID mechanism. © 2025, CC BY.

关键词： Crowdsourcing

Theory of Mixture-of-Experts for Mobile Edge Computing

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Li, Hongbo Duan, Lingjie Engineering Systems and Design Pillar Singapore University of Technology and Design Singapore

In mobile edge computing (MEC) networks, mobile users generate diverse machine learning tasks dynamically over time. These tasks are typically offloaded to the nearest available edge server, by considering communication and computational efficiency. However, its operation does not ensure that each server specializes in a specific type of tasks and leads to severe overfitting or catastrophic forgetting of previous tasks. To improve the continual learning (CL) performance of online tasks, we are the first to introduce mixture-of-experts (MoE) theory in MEC networks and save MEC operation from the increasing generalization error over time. Our MoE theory treats each MEC server as an expert and dynamically adapts to changes in server availability by considering data transfer and computation time. Unlike existing MoE models designed for offline tasks, ours is tailored for handling continuous streams of tasks in the MEC environment. We introduce an adaptive gating network in MEC to adaptively identify and route newly arrived tasks of unknown data distributions to available experts, enabling each expert to specialize in a specific type of tasks upon convergence. We derived the minimum number of experts required to match each task with a specialized, available expert. Our MoE approach consistently reduces the overall generalization error over time, unlike the traditional MEC approach. Interestingly, when the number of experts is sufficient to ensure convergence, adding more experts delays the convergence time and worsens the generalization error. Finally, we perform extensive experiments on real datasets in deep neural networks (DNNs) to verify our theoretical results. © 2024, CC BY.

关键词： Mobile edge computing

Online Learning from Strategic Human Feedback in LLM Fine-Tuning

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Hao, Shugang Duan, Lingjie Pillar of Engineering Systems and Design Singapore University of Technology and Design Singapore

Reinforcement learning from human feedback (RLHF) has become an essential step in fine-tuning large language models (LLMs) to align them with human preferences. However, human labelers are selfish and have diverse preferences. They may strategically misreport their online feedback to influence the system's aggregation towards their own preferences. Current practice simply averages labelers' feedback per time and fails to identify the most accurate human labeler, leading to linear regret O(T) for T time slots. To our best knowledge, we are the first to study online learning mechanisms against strategic human labelers in the LLM fine-tuning process. We formulate a new dynamic Bayesian game and dynamically adjust human labelers' weights in the preference aggregation, ensuring their truhtful feedback and sublinear regret O(T 1/2). Simulation results demonstrate our mechanism's great advantages over the existing benchmark schemes. Copyright © 2024, The Authors. All rights reserved.

关键词： Contrastive Learning

Entangling capabilities and unitary quantum games

学校读者我要写书评

暂无评论

Physical Review A 2024年第2期110卷 022413-022413页

作者： Rebecca Erbanni Antonios Varvitsiotis Dario Poletti Science Mathematics and Technology Cluster Engineering Systems Design Pillar Engineering Product Development Pillar National University of Singapore 117543 Singapore International Joint Research Unit

We consider a class of games between two competing players that take turns acting on the same many-body quantum register. Each player can perform unitary operations on the register, and after each one of them acts on the register the energy is measured. Player A aims to maximize the energy while player B to minimize it. This class of zero-sum games has a clear second mover advantage if both players can entangle the same portion of the register. We show, however, that if the first player can entangle a larger number of qubits than the second player (which we refer to as having quantum edge), then the second mover advantage can be significantly reduced. We study the game for different types of quantum edge of player A versus player B and for different sizes of the register, in particular, scenarios in which absolutely maximally entangled states cannot be achieved. In this case, we also study the effectiveness of using random unitaries. Last, we consider mixed initial preparations of the register, in which case the player with a quantum edge can rely on strategies stemming from the theory of ergotropy of quantum batteries.

关键词： Quantum entanglement Quantum information theory Economic games Quantum many-body systems

Algorithm design for Continual Learning in IoT Networks

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Hao, Shugang Duan, Lingjie Pillar of Engineering Systems and Design Singapore University of Technology and Design Singapore

Continual learning (CL) is a new online learning technique over sequentially generated streaming data from different tasks, aiming to maintain a small forgetting loss on previously-learned tasks. Existing work focuses on reducing the forgetting loss under a given task sequence. However, if similar tasks continuously appear to the end time, the forgetting loss is still huge on prior distinct tasks. In practical IoT networks, an autonomous vehicle to sample data and learn different tasks can route and alter the order of task pattern at increased travelling cost. To our best knowledge, we are the first to study how to opportunistically route the testing object and alter the task sequence in CL. We formulate a new optimization problem and prove it NP-hard. We propose a polynomial-time algorithm to achieve approximation ratios of 23 for underparameterized case and 32 +r1−T for overparameterized case, respectively. Simulation results verify our algorithm’s close-to-optimum performance. Copyright © 2024, The Authors. All rights reserved.

关键词： Approximation algorithms

Adaptive Decentralized Federated Learning in LEO Satellite Networks 24

学校读者我要写书评

暂无评论

Adaptive Decentralized Federated Learning in LEO Satellite N...

24th IEEE International Conference on Communication Technology, ICCT 2024

作者： Li, Dong Yan, Zhigang Hao, Yuyu Niyato, Dusit Quek, Tony Q. S. Macau University of Science and Technology School of Computer Science and Engineering Macau China Nanyang Technological University School of Computer Science and Engineering Singapore Singapore University of Technology and Design Information Systems Technology and Design Pillar Singapore

ISBN: (纸本)9798350363760

In Federated Learning (FL) for the low Earth orbit (LEO) satellite networks with parameters aggregated by a ground station, the communication overhead is a substantial concern. To circumvent this limitation and alleviate the latency within the FL for the satellite-ground communication, Decentralized Federated Learning (DFL) has appeared as an effective solution. However, considering the heterogeneity of satellites, the problem on how to efficiently leverage the limited resources available to enhance the model performance of DFL has drawn little attention, which motivates this work. Specifically, the proposed solution involves optimizing the number of local training rounds across diverse satellites with varying resource budgets to enhance the DFL performance by utilizing the resource more efficiently. Simulation results show that, the proposed scheme achieves a much better performance than the conventional schemes with fixed rounds of local training. © 2024 IEEE.

关键词： Federated learning

To Optimize Human-in-the-loop Learning in Repeated Routing Games

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Li, Hongbo Duan, Lingjie The Pillar of Engineering Systems and Design Singapore University of Technology and Design 487372 Singapore

Today navigation applications (e.g., Waze and Google Maps) enable human users to learn and share the latest traffic observations, yet such information sharing simply aids selfish users to predict and choose the shortest paths to jam each other. Prior routing game studies focus on myopic users in oversimplified one-shot scenarios to regulate selfish routing via information hiding or pricing mechanisms. For practical human-in-the-loop learning (HILL) in repeated routing games, we face non-myopic users of differential past observations and need new mechanisms (preferably non-monetary) to persuade users to adhere to the optimal path recommendations. We model the repeated routing game in a typical parallel transportation network, which generally contains one deterministic path and N stochastic paths. We first prove that no matter under the information sharing mechanism in use or the latest routing literature's hiding mechanism, the resultant price of anarchy (PoA) for measuring the efficiency loss from social optimum can approach infinity, telling arbitrarily poor exploration-exploitation tradeoff over time. Then we propose a novel user-differential probabilistic recommendation (UPR) mechanism to differentiate and randomize path recommendations for users with differential learning histories. We prove that our UPR mechanism ensures interim individual rationality for all users and significantly reduces PoA = ∞ to close-to-optimal (Equation presented), which cannot be further reduced by any other non-monetary mechanism. In addition to theoretical analysis, we conduct extensive experiments using real-world datasets to generalize our routing graphs and validate the close-to-optimal performance of UPR mechanism. Copyright © 2024, The Authors. All rights reserved.

关键词： Stochastic systems

Human-in-the-loop Learning for Dynamic Congestion Games

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Li, Hongbo Duan, Lingjie Pillar of Engineering Systems and Design Singapore University of Technology and Design Singapore487372 Singapore

Today mobile users learn and share their traffic observations via crowdsourcing platforms (e.g., Google Maps and Waze). Yet such platforms simply cater to selfish users’ myopic interests to recommend the shortest path, and do not encourage enough users to travel and learn other paths for future others. Prior studies focus on one-shot congestion games without considering users’ information learning, while our work studies how users learn and alter traffic conditions on stochastic paths in a human-in-the-loop manner. In a typical parallel routing network with one deterministic path and multiple stochastic paths, our analysis shows that the myopic routing policy (used by Google Maps and Waze) leads to severe under-exploration of stochastic paths. This results in a price of anarchy (PoA) greater than 2, as compared to the socially optimal policy achieved through optimal exploration-exploitation tradeoff in minimizing the long-term social cost. Besides, the myopic policy fails to ensure the correct learning convergence about users’ traffic hazard beliefs. To address this, we focus on informational (non-monetary) mechanisms as they are easier to implement than pricing. We first show that existing information-hiding mechanisms and deterministic path-recommendation mechanisms in Bayesian persuasion literature do not work with even PoA = ∞. Accordingly, we propose a new combined hiding and probabilistic recommendation (CHAR) mechanism to hide all information from a selected user group and provide state-dependent probabilistic recommendations to the other user group. Our CHAR mechanism successfully ensures PoA less than 45 , which cannot be further reduced by any other informational (non-monetary) mechanism. Besides the parallel network, we further extend our analysis and CHAR mechanism to more general linear path graphs with multiple intermediate nodes, and we prove that the PoA results remain unchanged. Additionally, we carry out experiments with real-world datasets to furt

关键词： Stochastic systems

Optimizing Multi-Device Wireless Power Transfer systems: A Novel Decoupling Strategy with Bipolar Coils 4

学校读者我要写书评

暂无评论

Optimizing Multi-Device Wireless Power Transfer Systems: A N...

4th International Conference on Computer Science, Electronic Information engineering and Intelligent Control Technology, CEI 2024

作者： Zhou, Yunzhi Hao, Min Shang, Chen Hu, Yaqi Xie, Shengli Kuzin, Victor School of Automation Guangdong University of Technology Guangzhou China Singapore University of Technology and Design Pillar of Information Systems Technology and Design Singapore Singapore Internet of Things Thrust Guangzhou China Russian Academy of Engineering Department of Geology and Engineering Moscow Russia

ISBN: (纸本)9798350378542

Wireless power transfer (WPT) technology provides key advantages over traditional charging methods, such as increased safety, convenience, and a broader charging range. However, single-receiver WPT systems lack resilience to transmitter coil failures and cannot simultaneously support multiple devices within the same charging area. Thus, the development of multi-load WPT systems is becoming increasingly urgent. A major challenge in multi-load WPT is cross-coupling, where unintended interactions between coils significantly reduce system efficiency and stability. Traditional methods to mitigate cross-coupling-such as adjusting coil placement, modifying coil shape, or adding complex compensation circuits-often involve intricate configurations that limit practical implementation or sacrifice efficiency. In contrast, this paper proposes a decoupling strategy using bipolar coils, offering a streamlined approach to suppress cross-coupling and enhance power stability in multi-load scenarios. Through this innovative solution, our study addresses limitations in current WPT systems, achieving improved reliability and output consistency without the complexity or efficiency trade-offs seen in traditional methods. © 2024 IEEE.

关键词： Inductive power transmission