A5028441149- Advanced Bandit Algorithms Research
- Distributed Control Multi-Agent Systems
- Transportation Planning and Optimization
- Game Theory and Applications
- Adaptive Dynamic Programming Control
- Mathematical and Theoretical Epidemiology and Ecology Models
- Vehicle Routing Optimization Methods
- Economic and Environmental Valuation
- Optimization and Variational Analysis
- Reinforcement Learning in Robotics
- Diffusion and Search Dynamics
- Neural Networks and Applications
- Statistical Mechanics and Entropy
- Metaheuristic Optimization Algorithms Research
- Advanced Thermodynamics and Statistical Mechanics
- Fuzzy Systems and Optimization
- Intracerebral and Subarachnoid Hemorrhage Research
- Facility Location and Emergency Management
- Stochastic Gradient Optimization Techniques
- Electric Vehicles and Infrastructure
- Antiplatelet Therapy and Cardiovascular Diseases
- Data Stream Mining Techniques
- Economic theories and models
- Acute Ischemic Stroke Management
- Sparse and Compressive Sensing Techniques
Purdue University West Lafayette
2021-2024
Tongji University
2020
Personalized prediction of the risk symptomatic intracerebral hemorrhage (sICH) after stroke thrombolysis is clinically useful. Machine-learning-based modeling may provide personalized sICH thrombolysis.We identified 2578 thrombolysis-treated ischemic patients between January 2013 and December 2016 from a multicenter database, where 70% were used to train models remaining 30% as nominal test sets. Another 136 consecutive tissue plasminogen-activated-treated 2017 our institute enrolled...
In this article, we focus on the stochastic generalized Nash equilibrium problem (SGNEP) which is an important and widely used model in many different fields. model, subject to certain global resource constraints, a set of self-interested players aims optimize their local objectives that depend own decisions others are influenced by some random factors. We propose distributed seeking algorithm partial-decision information setting based Douglas–Rachford operator splitting scheme, relaxes...
The aim of this paper is to find the distributed solution generalized Nash equilibrium problem (GNEP) for a group players that can communicate with each other over connected communication network. Each player tries minimize local objective function its own may depend on players' decisions, and collectively all decisions are subject some globally shared resource constraints. After reformulating optimization problems, we introduce notion network Lagrangian recast GNEP as zero finding properly...
A variety of practical problems can be modeled by the decision-making process in multi-player games where a group self-interested players aim at optimizing their own local objectives, while objectives depend on actions taken others. The gradient information each player, essential implementing algorithms for finding game solutions, is all too often unavailable. In this paper, we focus designing solution using bandit feedback, i.e., only available feedback player's disposal realized objective...
This paper investigates the equilibrium convergence properties of a proposed algorithm for potential games with continuous strategy spaces in presence feedback delays, main challenge multi-agent systems that compromises performance various optimization schemes. The is built upon an improved version accelerated gradient descent method. We extend it to decentralized scenario and equip delayed utilization scheme. By appropriately tuning step sizes studying interplay between delay functions...
Non-cooperative games serve as a powerful frame-work for capturing the interactions among self-interested players and have broad applicability in modeling wide range of practical scenarios, ranging from power management to drug delivery. Although most existing solution algorithms assume availability first-order information or full knowledge objectives others' action profiles, there are situations where only accessible at players' disposal is realized objective function values. In this paper,...
We explore a class of stochastic multiplayer games where each player in the game aims to optimize its objective under uncertainty and adheres some expectation constraints. The study employs an offline learning paradigm, leveraging pre-existing dataset containing auxiliary features. While prior research deterministic primarily explored vector-valued decisions, this work departs by considering function-valued decisions that incorporate features as input. leverage law large deviations degree...
Learning in multi-player games can model a large variety of practical scenarios, where each player seeks to optimize its own local objective function, which at the same time relies on actions taken by others. Motivated frequent absence first-order information such as partial gradients solving optimization problems and prevalence asynchronicity feedback delays multi-agent systems, we introduce bandit learning algorithm, integrates mirror descent, residual pseudo-gradient estimates,...
We consider a class of multi-agent optimization problems, where each agent has local objective function that depends on its own decision variables and the aggregate others, is willing to cooperate with other agents minimize sum objectives. After associating an auxiliary variable related estimates, we conduct primal decomposition globally coupled problem reformulate it so can be solved distributedly. Based Douglas-Rachford method, algorithm proposed which ensures exact convergence solution...
We consider the stochastic generalized Nash equilibrium problem (SGNEP) where a set of self-interested players, subject to certain global constraints, aim optimize their local objectives that depend on own decisions and others are influenced by some random factors. A distributed seeking algorithm is proposed based Douglas-Rachford operator splitting scheme, which only requires communications among neighbors. The scheme significantly relaxes assumptions co-coercivity contractiveness in...
This paper investigates the equilibrium convergence properties of a proposed algorithm for potential games with continuous strategy spaces in presence feedback delays, main challenge multi-agent systems that compromises performance various optimization schemes. The is built upon an improved version accelerated gradient descent method. We extend it to decentralized scenario and equip delayed utilization scheme. By appropriately tuning step sizes studying interplay between delay functions...
We consider a generalized Nash equilibrium problem (GNEP) for network of players. Each player tries to minimize local objective function subject some resource constraints where both the functions and depend on other players' decisions. By conducting equivalent transformations optimization problems introducing Lagrangian, we recast GNEP into an operator zero-finding problem. An algorithm is proposed based Douglas-Rachford method distributedly find solution. The requires milder conditions...
Learning in multi-player games can model a large variety of practical scenarios, where each player seeks to optimize its own local objective function, which at the same time relies on actions taken by others. Motivated frequent absence first-order information such as partial gradients solving optimization problems and prevalence asynchronicity feedback delays multi-agent systems, we introduce bandit learning algorithm, integrates mirror descent, residual pseudo-gradient estimates,...
Non-cooperative games serve as a powerful framework for capturing the interactions among self-interested players and have broad applicability in modeling wide range of practical scenarios, ranging from power management to drug delivery. Although most existing solution algorithms assume availability first-order information or full knowledge objectives others' action profiles, there are situations where only accessible at players' disposal is realized objective function values. In this paper,...
Non-cooperative games serve as a powerful framework for capturing the interactions among self-interested players and have broad applicability in modeling wide range of practical scenarios, ranging from power management to path planning self-driving vehicles. Although most existing solution algorithms assume availability first-order information or full knowledge objectives others' action profiles, there are situations where only accessible at players' disposal is realized objective function...
In stochastic Nash equilibrium problems (SNEPs), it is natural for players to be uncertain about their complex environments and have multi-dimensional unknown parameters in models. Among various SNEPs, this paper focuses on locally coupled network games where the objective of each rational player subject aggregate influence its neighbors. We propose a distributed learning algorithm based proximal-point iteration ordinary least-square estimator, repeatedly updates local estimates neighboring...
In this paper, we focus on the stochastic generalized Nash equilibrium problem (SGNEP) which is an important and widely-used model in many different fields. model, subject to certain global resource constraints, a set of self-interested players aim optimize their local objectives that depend own decisions others are influenced by some random factors. We propose distributed seeking algorithm partial-decision information setting based Douglas-Rachford operator splitting scheme, relaxes...
We consider a class of multi-agent optimization problems, where each agent has local objective function that depends on its own decision variables and the aggregate others, is willing to cooperate with other agents minimize sum objectives. After associating an auxiliary variable related estimates, we conduct primal decomposition globally coupled problem reformulate it so can be solved distributedly. Based Douglas-Rachford method, algorithm proposed which ensures exact convergence solution...