陳柏安

Optimal Freight Allocation for Freight Forwarders under Stochastic Demand and Two Airline Discount Mechanisms

運輸與物流管理學系物流管理・共同指導：姚銘忠

貨運代理商、隨機規劃、馬可夫決策過程、深度Q網路、全量折扣、Freight Forwarder、Stochastic Programming、Markov Decision Process、Deep Q Network、All-Unit Discount

Cost-Aware Text-to-Knowledge Graph Construction with Large Language Models: An Inference-Time Evidence Selection Approach

資訊管理研究所・共同指導：張宏輔

大型語言模型、文本到知識圖譜、知識圖譜建構、證據選擇、成本感知推論、Teacher–Student Learning、Large Language Models、ext-to-Knowledge Graph、Knowledge Graph Construction、Evidence Selection、Cost-Aware Inference、eacher–Student Learning

Large Language Model Safety Alignment Integrating Parity Constraints: Experiments and Analysis of the KTOSP Algorithm

資訊管理研究所

大型語言模型、安全對齊、康納曼-特費斯基最佳化、安全平權、過度擬合、對齊稅、Large Language Models、Safety Alignment、Kahneman-Tversky Optimization、Safety Parity、Lexical Overfitting、Alignment Tax

Learning Time-to-Value Functions in (Atomic Splittable) Congestion Control

資訊管理研究所

原子可分割壅塞控制、時間價值函數、Competitive Demand Learning、動態定價、流量生成、串並聯網路、Atomic splittable congestion control、time-to-value function、dynamic pricing、flow generation、series-parallel network

Meta-Learning for Convergence to Equilibria and the Price of Anarchy in General-Sum Markov Games

資訊管理研究所

元學習、一般和馬可夫賽局、納許均衡、相關均衡、無中控代價、雙線性鞍點問題、Meta-learning、General-sum Markov games、Nash equilibrium、Correlated equilibrium、Price of anarchy、Bilinear saddle-point problem

Multiagent Restarted Q-Learning with Upper Confidence Bounds for Non-Stationary Markov Games: Congestion Games as an Example

資訊管理研究所

馬可夫賽局、非平穩環境、多代理人、擁塞賽局、Markov Games、Non-Stationary Environment、Multiagent、Congestion Games

Enhancing Symmetric Crystal Generative Models via Refinement and Conditional Training

資訊管理研究所

材料設計、生成模型、對稱晶體生成、魏科夫位置、變換器、Materials Design、Generative Model、Symmetric Crystal Generation、Wyckoff Position、Transformer

Fairness and Safety in Language Models: Conditional Self-supervised Contrastive Learning for Bias Mitigation and Toxicity Assessment of Diffusion Models

資訊管理研究所

文本分類、對比學習、公平性、語言擴散模型、text classification、contrastive learning、fairness、diffusion language models

Structure-Aware Robust Counterfactual Explanations via Conditional Gaussian Network Classifiers

資訊管理研究所・共同指導：兪在元

反事實解釋、可解釋人工智慧、對抗式框架、穩健最佳化、條件高斯網路分類器、Counterfactual Explanation、Explainable Artificial Intelligence、Adversarial Framework、Robust Optimization、Conditional Gaussian Network Classifier

LLM-Based SQL Generation: Prompting, Self-Refinement,and Adaptive Weighted Majority Voting

資訊管理研究所・共同指導：張宏輔

自然語言轉結構式查詢語言、大型語言模型、加權多數決演算法、自我精鍊機制、Text-to-SQL、Large Language Models (LLMs)、Weighted Majority Algorithm (WMA)、Self-Refinement

Resolving Multi-Agent Competition and Cooperation in Logistic Tasks via Prompting Large Language Models

資訊管理研究所

大型語言模型、多代理人、虛擬助理、提示工程、生成式預訓練轉換、large language model、multi-agent、prompt engineering、embodied agent、GPT

Online Firefighting on Trees with Predictions

資訊管理研究所・共同指導：劉向瑄

擴增學習法、線上演算法、消防員問題、Firefighter Problem、Learning Augmented、Robustness、Greedy Algorithms

Conditional Supervised Contrastive Learning for Fair and Robust Non-English Text Classification

資訊管理研究所

文本分類、對比學習、公平性、穩健性、Text classification、Contrastive learning、Fairness、Robustness

Randomization and Distribution-Based Relaxation in Robust Counterfactual Explanations for Decision Tree Models

資訊管理研究所

機器學習、可解釋性人工智慧、反事實解釋、穩健最佳化、決策樹模型、machine learning、explainable AI、counterfactual explanation、robust optimization、decision tree

Playing General-Sum Markov Games with Generic and Specific No-Regret Algorithms

資訊管理研究所

馬可夫賽局、一般和賽局、無悔演算法、Markov Game、General-sum game、No-regret algorithm

Adversarial Contrastive Learning for Fair Clustering

資訊管理研究所

深度學習分群法、對比式學習、對比式分群法、對抗式學習、公平性、Deep Clustering、Contrastive Learning、Contrastive Clustering、Adversarial Learning、Fairness

Robust Multi-Maker Market Making via Distributed Gradient-Based Algorithms

資訊管理研究所

做市商交易、馬可夫博弈、樂觀梯度下降上升、最後迭代收斂、Market Making、Markov Games、Optimistic Gradient Descent Ascent、Last-iterate Convergence

Correlation-Based Self-Supervised Learning for Fair High-Dimensional Data Classification

資訊管理研究所

偏見與公平性、公平表示學習、自我監督學習、基於相關性的去偏見方法、Bias and Fairness、Fair Representation Learning、Self-Supervised Learning、Correlation- Based Debiasing

Model-Agnostic Training with Sample Selection for Fair and Robust Learning

資訊管理研究所

可信任人工智慧、公平與穩健訓練、無關模型訓練、不公平消弭、穩健收斂、trustworthy AI、fair and robust training、model-agnostic、unfairness mitigation、convergence to robustness

Post-Processing Techniques for Machine Learning Debias and Explainability

資訊管理研究所

機器學習、偏誤與公平、可解釋AI、注意力機制、SHAP、Machine Learning、Bias and Fairness、Explainable AI、Attention Mechanism

Multi-Agent Pathfinding via Correlated Q-Learning

資訊管理研究所

多代理人路徑搜尋、強化式學習、Q-Learning、相關性均衡Q-Learning、無悔學習、相關性均衡、粗的相關性均衡、Multi-Agent Pathfinding、Reinforcement Learning、Correlated Q-Learning、No-Regret Learning、Correlated Equilibrium、Coarse Correlated Equilibrium

Adversarial Learning with Optimism for Debias in Machine Learning

資訊管理研究所

偏見與公平、對抗式學習、對抗式消除偏見、樂觀式優化器、收斂速度、Bias and Fairness、Adversarial Learning、Adversarial Debiasing、Optimistic Optimizer、Convergence Rate

An Improved Privacy-Preserving Matching Mechanism for Multi-Sided Online Advertising

資訊管理研究所

線上廣告、機制設計、激勵相容、無悔演算法、收益最大化、Online Advertising、Mechanism Design、Incentive Compatible、No-regret learning、Gain from trade maximization

Scheduling Algorithms for Hospital Surgery Department Debottlenecking

資訊管理研究所

手術排程、強化學習、線性規劃鬆弛、Surgery scheduling、Reinforcement learning、Linear programming relaxation

Under-Sampling and Over-Sampling for Debias in Machine Learning

資訊管理研究所

機器學習偏見、組合採樣增強、AI 公平性、集成學習、SHAP、Machine Learning Bias、Combination Sampling Augmentation、AI Fairness、Ensemble Learning

Multiagent Learning Algorithms for Competitive Opinion Optimization

資訊管理研究所

競爭意見最佳化、樂觀梯度下降上升、多代理人學習、無悔學習、極小極大均衡、Competitive Opinion Optimization、Optimistic Gradient Descent Ascent、Multiagent Learning、No-regret Learning、Minimax Equilibrium

Multi-Agent Pathfinding via A* Enhanced Hierarchical Actor-Critic Reinforcement Learning

管理學院資訊管理學程

多代理人路徑搜尋、分層強化學習、A*演算法、目標導向、優先經驗回放、MAPF、HAC、RL、Goal-oriented、A* algorithm、PER、APE-X

Distributed Algorithms for Finding Linear Arrow-Debreu Market Equilibria

資訊管理研究所

Arrow-Debreu市場均衡、凸最佳化問題、極小化極大問題、分散式演算法、樂觀式梯度下降與上升演算法、Arrow-Debreu market equilibria、Convex program、Minimax program、Distributed algorithm、Optimistic gradient descent ascent

Opinion Optimization on Social Networks via Mixed Integer Programming and Relaxations with Rounding

管理學院資訊管理學程

意見形成的賽局、意見最佳化、混合整數規劃、Opinion­ Forming Games、Opinion Optimization、Mixed­ Integer Programming

Combinatorial Optimization in Social Networks via Machine Learning

資訊管理研究所

影響力最大化、意見最大化、機器學習、分類器、社群理論、Influence maximization、Opinion maximization、Machine learning、Classifier、Social theory

Reinforcement Learning for Job Scheduling

管理學院資訊管理學程

強化學習、生產排程、Reinforcement Learning、Job Scheduling

Weak Equilibria and Approximate Nash Equilibria in Scheduling Games with Multi­-Job Players

資訊管理研究所

排程賽局、排程政策、弱均衡、納許均衡、無中控代價、改善率、Job Scheduling Games、Coordination Mechanisms、Weak Equilibrium、Nash Equilibrium、Price of Anarchy、Improvement Ratio

Opinion Maximization: Mixed Integer Programming and LP Rounding for Directed Acyclic Graphs

資訊管理研究所

意見最大化、混和整數規畫、線性規劃、有向無環圖、有向樹、Opinion maximization、Mixed Integer Programming、Linear Program­ming Rounding、Directed acyclic graph、Directed tree

Distributed Dynamic Multi-Agent Task Assignment and Path Finding via Reinforcement Learning

資訊管理研究所

強化學習、A3C、任務指派、多代理人路徑規劃、TAPF、RL、Task-Assignment、MAPF

The Price of Anarchy for Two Randomized Local Policies in Scheduling Games with Single-Job and Multi-Job Players: Numerical Computation

資訊管理研究所

排程賽局、隨機策略、本機政策、多工賽局、單工賽局、Scheduling Game、Local Policy、Price of Anarchy、Rand、Externality

Firefighter Problem with Spreading Vaccination for Maximizing the Number of Saved Nodes: Exact and Approximation Algorithms

資訊管理研究所

整數規劃、線性規劃的鬆弛、疫苗注射策略、精準演算法、近似演算法、Integer Linear Programming、LP Relaxation、Vaccination Strategy Problem、Exact Algorithm、Approximation Algorithm

Dynamic Pricing through Stochastic Online Learning

資訊管理研究所

動態定價、隨機線上學習、角子老虎機問題、Dynamic Pricing、Stochastic Online Learning、Multi-Armed Bandit

Stateless Q-Learning for Multi-Agent Collision-Free Path Finding

資訊管理研究所

多代理人路徑搜尋、增強式學習、網路流、分散式計算、Multi-agent path finding、Reinforcement Learning、Network flows、Distributed computing

Selling User Data to No-Regret Advertisers: Dynamic Multi-Item Auctions

資訊管理研究所

動態拍賣、線上廣告、無悔學習、機制設計、dynamic auctions、online advertising、no-regret learning、mechanism design

One-Team Multiagent Collision-Free Path Finding and Task Assignment via Algorithms for Finding Network Flows

資訊管理研究所

網絡流、多代理人路徑與避免碰撞的任務指派、Network flow、Multi-agent path finding and task assignment

Randomized Local Policies for Scheduling Games with Multi-jobs Players

資訊管理研究所

賽局理論、排程理論、多工玩家、奈許均衡、最壞均衡與最佳解比、完工時間、Game Theory、Scheduling、Multi-Job、Nash equilibrium、Price of anarchy、Completion time

Market Equilibria and Games: An Algorithmic Approach

資訊管理研究所

賽局理論、市場均衡、分散式演算法、Market Equilibria、Alternating Minimization、Price of Anarchy

Opinion Formation Games in Social Networks

資訊管理研究所

意見、賽局理論、納許均衡、Opinions、Games、Nash Equilibrium、Price of Anarchy

A Study of the Learning Effect of Teaching Object-Oriented Programming Course with Programmed Instruction

資訊管理研究所・共同指導：楊千

程式設計、圖型化介面、物件導向、強化理論、編序教學、Object-Oriented、Programmed Instruction、Java、Reinforcement Theory、Graphic User Interface