具有未知动态的线性系统二人零和博弈问题在线学习方案

针对具有未知动态线性系统的二人零和博弈问题,本文提出了一种新的基于单环迭代方法的在线学习方案.为保证单环迭代方法的收敛性,给出了一种新的分析方法.在系统内部矩阵A,控制输入矩阵B以及干扰输入矩阵D均未知的情况下,通过在线迭代策略,同步得到了博弈代数黎卡提方程的近似解,以及控制和干扰策略.仿真结果表明了所提方法的有效性.     

Characterizing land cover dynamics using multi-temporal imagery

An analysis of land cover changes was performed using a time-series of five SPOT HRV images for an area of the State of Rond°onia (western Brazilian Amazon) from 1986 to 1992. The total deforested area and the fraction of land abandoned to secondary vegetation were determined by means of image classification and Geographical Information System (GIS) techniques. Areas deforested by 1986 were traced throughout the period to estimate the fraction of land remaining continuously in the secondary vegetation category, possibly forming older secondary vegetation.     

Uncoupling Isaacs equations in two-player nonzero-sum differential games. Parental conflict over care as an example

We consider a two-player nonzero-sum differential game in mixed strategies between two players who have, at each instant of time, two pure strategies to mix between. The search for a Nash equilibrium in state feedback leads to a pair of uncoupled first order PDEs. We give an example in Behavioral Ecology.     

基于贝叶斯博弈模型的无线多跳网络激励机制

节点智能性的提高使无线多跳网络的应用更为多样化,但也使网络的安全问题变得更为突出。为了有效控制自私节点和恶意节点给网络带来的不良影响,在博弈论的基础上结合跨层优化思想,构建了物理层与链路层部分信息共享的贝叶斯博弈模型。利用贝叶斯博弈理论对节点的交互信息构建策略空间并进行推导分析,形成有效激励节点合作的互监督机制。最后,分别通过典型案例与计算机仿真实验验证了该网络模型的可行性和公平性。     

环境污染博弈问题的系统动力学模型*

用系统动力学建立环境污染管理问题中政府管理部门与生产排污企业之间的一个混合战略重复博弈模型,考虑信息延迟和环境的污染量净化问题进行模型的仿真。结果表明,环境污染问题中政府管理部门与生产排污企业之间博弈的Nash均衡实际上很难达到稳定,尤其是在考虑信息延迟的情况下。最后提出一种简单可行的双重惩罚策略,既可以促使博弈双方尽快达到Nash均衡,也能有效地改善环境污染问题。     

RTS game strategy evaluation using extreme learning machine

The fundamental game of real-time strategy (RTS) is collecting resources to build an army with military units to kill and destroy enemy units. In this research, an extreme learning machine (ELM) model is proposed for RTS game strategy evaluation. Due to the complicated game rules and numerous playable items, the commonly used tree-based decision models become complex, sometimes even unmanageable. Since complex interactions exist among unit types, the weighted average model usually cannot be well used to compute the combined power of unit groups, which results in misleading unit generation strategy. Fuzzy measures and integrals are often used to handle interactions among attributes, but they cannot handle the predefined unit production sequence which is strictly required in RTS games. In this paper, an ELM model is trained based on real data to obtain the combined power of units in different types. Both the unit interactions and the production sequence can be implicitly and simultaneously handled by this model. Warcraft III battle data from real players are collected and used in our experiments. Experimental results show that ELM is fast and effective in evaluating the unit generation strategies.     

Chaos in game dynamics

Two examples demonstrate the possibility of extremely complicated non-convergent behavior in evolutionary game dynamics. For the Taylor-Jonker flow, the stable orbits for three strategies were investigated by Zeeman. Chaos does not occur with three strategies. This papers presents numerical evidence that chaotic dynamics on a strange attractor does occur with four strategies. Thus phenomenon is closely related to known examples of complicated behavior in Lotka-Volterra ecological models.I would like to thank John Doyle for raising the question of chaos in game dynamics and for pointing out the review article of Baumol and Benhabib (1989). Steve Frank called my attention to Smale (1976) and Vandermeer (1991). Thanks also to Brad Armendt, J.H. Nachbar, Stergios Skapardas, Eliott Sober and an anonymous referee for comments on an earlier version of this paper.     

模糊网络博弈的合作联盟优化对策研究

模糊网络博弈主要关注如何将联盟收益分配给合作联盟的每个参与者,其广义模糊博弈解同时引入参与度和调整系数,不仅实现了参与者以部分资源参与合作联盟的愿望,而且满足了保留部分收益值用于联盟再发展的需求.本文作为模糊博弈模型的后续深入研究,对模糊网络博弈的解进行拓展,提出广义模糊核心解、广义模糊谈判集解的概念,并证明当满足超可加性的前提下,模糊网络博弈的广义模糊核心解与其广义模糊谈判集解具有等价关系,并刻画了模糊网络博弈广义核心解的非空性,算例分析结果表明合作联盟广义分配方案的存在性,为合作联盟优化对策提供服务.     

Dynamics in network interaction games

We study the convergence times of dynamics in games involving graphical relationships of players. Our model of interaction games generalizes a variety of recently studied games in game theory and distributed computing. In a local interaction games each agent is a node embedded in a graph and plays the same 2-player game with each neighbor. He can choose his strategy only once and must apply his choice in each 2-player game he is involved in. This represents a fundamental model of decision making with local interaction and distributed control. Furthermore, we introduce a generalization called 2-type interaction games, in which one 2-player game is played on edges and possibly another game is played on non-edges. For the popular case with symmetric 2 ×?2 games, we show that several dynamics converge to a pure Nash equilibrium in polynomial time. This includes arbitrary sequential better-response dynamics, as well as concurrent dynamics resulting from a distributed protocol that does not rely on global knowledge. We supplement these results with an experimental comparison of sequential and concurrent dynamics.     

使用长短期记忆网络预测NBA比赛胜负

深度学习与机器学习的方法已广泛应用于NBA(美国篮球职篮联赛)的比赛胜负的预测中,然而过去的方法未对过去几场比赛的数据进行建模,忽略了比赛双方近期状态的有效表示.为了解决这个问题,提出了基于长短期记忆(LSTM)网络的方法对NBA常规赛的比赛胜负进行预测.该方法分别以比赛中的两支球队过去几场比赛的数据作为LSTM的输入,以该场比赛结果作为输出,训练能够预测比赛胜负的模型.本质上是使用球队在该赛季的历史数据的平均值作为该球队的实力,以近几场比赛的数据序列作为该球队状态的体现.在实验中比较了其他几种预测NBA比赛胜负的方法(支持向量机、卷积神经网络、逻辑回归模型等方法),数据来自2014—2019年间的5个赛季的NBA常规赛数据.结果表明,模型的预测准确率达到(69.09％),高于其他几种模型.     

WSN中利用贝叶斯博弈的数据传递策略

在无线传感器网络中,数据的传递策略对网络的能量损耗具有重要的影响,为此,提出了一个基于贝叶斯博弈的数据传递模型。在该模型中网络节点为了获取最大的收益,在考虑自身能量水平的基础上,适当的调整发送/转发的数据量。当节点发送/转发的数据满足一定条件时,网络存在均衡状态。仿真结果表明,该基于博弈论的数据传递策略在均衡状态下能够明显降低能量损耗,延长网络的使用寿命。     

使用长短期记忆网络预测NBA比赛胜负

深度学习与机器学习的方法已广泛应用于NBA(美国篮球职篮联赛)的比赛胜负的预测中,然而过去的方法未对过去几场比赛的数据进行建模,忽略了比赛双方近期状态的有效表示.为了解决这个问题,提出了基于长短期记忆(LSTM)网络的方法对NBA常规赛的比赛胜负进行预测.该方法分别以比赛中的两支球队过去几场比赛的数据作为LSTM的输入,以该场比赛结果作为输出,训练能够预测比赛胜负的模型.本质上是使用球队在该赛季的历史数据的平均值作为该球队的实力,以近几场比赛的数据序列作为该球队状态的体现.在实验中比较了其他几种预测NBA比赛胜负的方法(支持向量机、卷积神经网络、逻辑回归模型等方法),数据来自2014—2019年间的5个赛季的NBA常规赛数据.结果表明,模型的预测准确率达到(69.09％),高于其他几种模型.     

基于模糊静态贝叶斯博弈的网络主动防御策略选取

针对网络攻防过程中攻防双方存在无法完全获知对方信息以及无法对双方损益作出准确判定的问题,提出了一种基于模糊贝叶斯博弈模型的网络最优防御策略选取方法,结合网络攻防对抗双方信息不完全性的特点构建了模糊静态贝叶斯博弈模型.在此基础上引入三角模糊数描述攻防双方的效用函数,采用基于模糊概率及均值的方法获得清晰效用函数值,进而通过求解模型的纳什均衡获得最优防御策略;最后给出了最优防御策略选取流程.理论分析和仿真实验表明,该方法能够对不同攻击行为概率作出有效预测,为主动防御提供决策支持.     

Concurrent imitation dynamics in congestion games

Imitating successful behavior is a natural and frequently applied approach when facing complex decision problems. In this paper, we design protocols for distributed latency minimization in atomic congestion games based on imitation. We propose to study concurrent dynamics that emerge when each agent samples another agent and possibly imitates this agent’s strategy if the anticipated latency gain is sufficiently large. Our focus is on convergence properties. We show convergence in a monotonic fashion to stable states, in which none of the agents can improve their latency by imitating others. As our main result, we show rapid convergence to approximate equilibria, in which only a small fraction of agents sustains a latency significantly above or below average. Imitation dynamics behave like an FPTAS, and the convergence time depends only logarithmically on the number of agents. Imitation processes cannot discover unused strategies, and strategies may become extinct with non-zero probability. For singleton games we show that the probability of this event occurring is negligible. Additionally, we prove that the social cost of a stable state reached by our dynamics is not much worse than an optimal state in singleton games with linear latency functions. We concentrate on the case of symmetric network congestion games, but our results do not use the network structure and continue to hold accordingly for general symmetric games. They even apply to asymmetric games when agents sample within the set of agents with the same strategy space. Finally, we discuss how the protocol can be extended such that, in the long run, dynamics converge to a pure Nash equilibrium.     

Power in threshold network flow games

Preference aggregation is used in a variety of multiagent applications, and as a result, voting theory has become an important topic in multiagent system research. However, power indices (which reflect how much “real power” a voter has in a weighted voting system) have received relatively little attention, although they have long been studied in political science and economics. We consider a particular multiagent domain, a threshold network flow game. Agents control the edges of a graph; a coalition wins if it can send a flow that exceeds a given threshold from a source vertex to a target vertex. The relative power of each edge/agent reflects its significance in enabling such a flow, and in real-world networks could be used, for example, to allocate resources for maintaining parts of the network. We examine the computational complexity of calculating two prominent power indices, the Banzhaf index and the Shapley-Shubik index, in this network flow domain. We also consider the complexity of calculating the core in this domain. The core can be used to allocate, in a stable manner, the gains of the coalition that is established. We show that calculating the Shapley-Shubik index in this network flow domain is NP-hard, and that calculating the Banzhaf index is #P-complete. Despite these negative results, we show that for some restricted network flow domains there exists a polynomial algorithm for calculating agents’ Banzhaf power indices. We also show that computing the core in this game can be performed in polynomial time.     

Interactive strategy sets in multiple payoff games

In game theory, it is usually assumed that each player has only one payoff function and the strategy set of the game is composed of the topological product of individual players’ strategy sets. In real business and system design or control problems, however, players’ strategy sets may be interactive and each player may have more than one payoff function. This paper, investigates the more general situation of multiple payoff and multiple person games in a normal form. In this paper, each player has several payoff functions which are dominated by certain convex cones, and the feasible strategy set of each player may be interactive with those of the other players. This new model is applied to a classical example without requiring variational and quasi-variational inequalities, or point-to-set mappings.     

Characterizing files in the modern Gnutella network

The Internet has witnessed an explosive increase in the popularity of Peer-to-Peer (P2P) file-sharing applications during the past few years. As these applications become more popular, it becomes increasingly important to characterize their behavior in order to improve their performance and quantify their impact on the network. In this paper, we present a measurement study on characteristics of available files in the modern Gnutella system. We develop two new methodologies to capture accurate snapshots of available files in a large-scale P2P system. These methodologies were implemented in a parallel crawler that captures the entire overlay topology of the system where each peer in the overlay is annotated with its available files. We have captured more than 50 snapshots of the Gnutella system that span over 1 year period. Using these snapshots, we conduct three types of analysis on available files: (1) Static analysis, (2) Topological analysis, and (3) dynamic analysis. Our results reveal several interesting properties of available files in Gnutella that can be leveraged to improve the design and evaluation of P2P file-sharing applications. This paper extends and supplants the earlier version of this paper presented at MMCN 2006 [1]. This material is based upon work supported by the National Science Foundation (NSF) under Grant No. Nets-NBD-0627202, CAREER Award CNS-0448639, and an unrestricted gift from Cisco Systems. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF or Cisco.     

基于邻域选择策略的图卷积网络模型

邻域的组成对于基于空间域的图卷积网络（GCN）模型有至关重要的作用。针对模型中节点邻域排序未考虑结构影响力的问题,提出了一种新的邻域选择策略,从而得到改进的GCN模型。首先,为每个节点收集结构重要的邻域并进行层级选择得到核心邻域;然后,将节点及其核心邻域的特征组成有序的矩阵形式;最后,送入深度卷积神经网络（CNN）进行半监督学习。节点分类任务的实验结果表明,该模型在Cora、Citeseer和Pubmed引文网络数据集中的节点分类准确性均优于基于经典图嵌入的节点分类模型以及四种先进的GCN模型。作为一种基于空间域的GCN,该模型能有效运用于大规模网络的学习任务。