Optimal mixed block withholding attacks based on reinforcement learning

The vulnerabilities in cryptographic currencies facilitate the adversarial attacks. Therefore, the attackers have incentives to increase their rewards by strategic behaviors. Block withholding attacks (BWH) are such behaviors that attackers withhold blocks in the target pools to subvert the blockchain ecosystem. Furthermore, BWH attacks may dwarf the countermeasures by combining with selfish mining attacks or other strategic behaviors, for example, fork after withholding (FAW) attacks and power adaptive withholding (PAW) attacks. That is, the attackers may be intelligent enough such that they can dynamically gear their behaviors to optimal attacking strategies. In this paper, we propose mixed-BWH attacks with respect to intelligent attackers, who leverage reinforcement learning to pin down optimal strategic behaviors to maximize their rewards. More specifically, the intelligent attackers strategically toggle among BWH, FAW, and PAW attacks. Their main target is to fine-tune the optimal behaviors, which incur maximal rewards. The attackers pinpoint the optimal attacking actions with reinforcement learning, which is formalized into a Markov decision process. The simulation results show that the rewards of the mixed strategy are much higher than that of honest strategy for the attackers. Therefore, the attackers have enough incentives to adopt the mixed strategy.     

具有切换拓扑的多智能体系统协同控制的差分隐私保护

网络信息技术的不断发展与普及使得各类数据的发布采集变得方便与便捷, 但数据的直接发布势必会造 成个网络信息的泄露和敏感信息的失密, 因此敏感信息的保护成为了各行各业关注的问题. 本文研究了基于固定拓 扑和切换拓扑的多智能体系统协同控制的差分隐私保护问题, 将差分隐私算法与传统平均一致性算法结合, 提出了 具有隐私保护的协同控制算法, 分析了隐私保护算法对分布式协同控制闭环系统稳定性的影响. 基于所提算法, 应 用矩阵论和概率统计对隐私保护协同控制算法的收敛性和隐私性进行理论分析, 该算法可以保护智能个体的数据 隐私, 同时可以使得系统运动实现均方一致. 在系统拓扑结构动态变化的情况下, 本文对该算法的收敛性和隐私性 进行理论分析, 讨论了切换拓扑对隐私保护的影响. 最后的仿真示例验证了理论结果的正确性.     

基于映射字典学习的跨模态哈希检索

针对网络上出现越来越多的多模态数据,如何在海量数据中检索不同模态的数据成为一个新的挑战.哈希方法把数据映射到Hamming空间,大大降低了计算复杂度,为海量数据的跨模态检索提供了一条有效的路径.然而,大部分现存方法生成的哈希码不包含任何语义信息,从而导致算法性能的下降.为了解决这个问题,本文提出一种基于映射字典学习的跨模态哈希检索算法.首先,利用映射字典学习一个共享语义子空间,在子空间保持数据模态间的相似性.然后,提出一种高效的迭代优化算法得到哈希函数,但是可以证明问题的解并不是唯一的.因此,本文提出通过学习一个正交旋转矩阵最小化量化误差,得到性能更好的哈希函数.最后,在两个公开数据集上的实验结果说明了该算法优于其他现存方法.     

基于囚徒困境策略的改进HK网络上的合作博弈

为模拟现实世界的合作行为,本文在HK网络模型基础上提出了一种具有高聚类幂律可调性质的新的网络模型,并分析了囚徒困境博弈在此网络上的演化。通过仿真实验,研究了该网络的高聚类特性对合作行为的影响。大量实验表明,网络的高聚类特性可以极大促进合作现象的涌现。同时研究也发现,随着诱惑参数的变大,合作水平也会随之下降,但幅度不大。总之,该演化博弈模型可以促进合作现象的涌现并抵御背叛策略的传播。     

Distributed adaptive tracking control for high-order multi-agent systems with unknown dynamics

In this paper, we investigate the adaptive tracking problem of high-order multi-agent systems with unknown parameters and unknown nonlinear functions. Under the assumption that the leader is the root of a spanning tree, a distributed adaptive controller with tuning function is constructed recursively based on backstepping design method. The designed controller can guarantee that the tracking errors and the parameter errors eventually converge to an arbitrarily small compact set by choosing design parameters. A simulation example demonstrates the effectiveness of the design scheme.     

基于时空序列搜索的最优个性化旅游规划

随着人们对高效、个性化服务的日益关注,其对旅游规划的要求也越来越高。然而,目前制定一个完美的、个性化的旅游规划很困难。一个合理的旅游规划应该包含以下特征:1)基于游客兴趣爱好的个性化推荐;2)覆盖尽可能多的兴趣景点;3)走最少的路程,花费最少的时间。由于每次旅游包含多个景点,因此该问题变成了每个景点都附有权值的时空序列优化问题。为了实现最优个性化旅游规划,首先,针对每个城市为每位游客提供一系列带有权值的景点;然后,综合考虑距离、权值、旅游时间和景点特色等因素,为游客推荐一个最优旅游序列。为此,提出了4种有效算法,使游客在旅游时间范围限制内走最少的路程来游览最想游览的景点。在社交网络的真实数据集上,通过实验验证了所提算法的有效性。     

求解人力资源分配问题的多目标微粒群优化算法*

提出一种改进的多目标微粒群优化算法来求解人力资源分配问题.通过对种群进行正交初始化,保证了个体在整个可行解空间上的均匀分散,使得算法能够在整个可行解空间上进行均匀搜索;通过基于网格技术的外部存档非劣解删选策略,有效地保留了逼近Pareto前沿的非劣解;引入一种广义的学习策略来提升粒子向Pareto前沿收敛的概率.实验结...     

New insight into the mechanism of enhanced crystallization of PLA in PLLA/PDLA mixture

Poly(lactic acid) (PLA) is a bio‐based and compostable polymer that has quickly developed into a competitive material, but the control of crystallinity is a bottleneck in extended utilization. The crystallization of PLA has been a rich topic because of the existence of two enantiomeric forms of poly(L‐lactic acid) (PLLA) and poly(d ‐lactic acid) (PDLA) can form stereocomplex (SC) crystal with high melting point that can be used to control the crystallization behaviors. The SC crystal was regarded as an effective nucleating agent for promoting the crystallization rate and crystallinity of PLA. We investigated cold crystallization of PLLA/PDLA (1:1) mixture with in situ WAXS measurements and found that the homo‐crystal of PLA formed earlier than the SC‐crystal in the mixture within the measured temperature range, which is different from the melting crystallization. The final crystalline structures are in correspondence with the melting and cold crystallization temperature, and the transition of homo‐PLA (δ to α) is not altered by the crystallization procedure. The SC‐crystal can be detected in both cold and melting crystallization of the mixture at the temperatures lower than 150 °C, which is conflict with the reported results. A new crystallization mechanism of the mixture was proposed to understand the crystallization behaviors in PLLA/PDLA mixtures. © 2017 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2018 , 135, 45663.     

Synthesis of novel polymer brushes of poly(acrylonitrile-<Emphasis Type="Italic">g</Emphasis>-<Emphasis Type="Italic">N,N</Emphasis>ʹ-dimethylaminoethyl methacrylate) by nitrile modification

Functional polymer brushes of poly(acrylonitrile-g-N,N?-dimethylaminoethyl methacrylate) were efficiently synthesized by a novel approach of combining Cu(0)-mediated controlled/“living” radical polymerization and nitrile click chemistry. The poly(acrylonitrile-g-N,N?-dimethylaminoethyl methacrylate) as a promising material exhibited excellent hydrophile–lipohile balance in the process of self-assembly and could autonomously develop the orderly structure micelle in N,N?-dimethylformamide/water mixture solvent for the potential application of a new drug delivery carrier. In the process of self-assembly, polyacrylonitrile acted as a backbone of the functional polymer brushes due to its hydrophobic feature and poly(N,N?-dimethylaminoethyl methacrylate) as the branch of functional polymer brushes due to its hydrophilic characteristic, which both were prepared by Cu(0)-mediated controlled/“living” radical polymerization with ethyl-bromoisobutyrate as initiator. Poly(N,N?-dimethylaminoethyl methacrylate) containing azide end group was synthesized by substitution reaction of poly(N,N?-dimethylaminoethyl methacrylate) containing bromine end groups with sodium azide in N,N?-dimethylformamide. The click reaction between the nitrile of polyacrylonitrile and the azide group of poly(N,N?-dimethylaminoethyl methacrylate) was carried out under ammonium chloride as catalyst in N,N?-dimethylformamide. The polymer was further confirmed by GPC, FTIR, ¹H NMR and TGA. Meanwhile, the micelles with different morphologies were observed by TEM, and the particle diameter distribution of self-assembled micelle from the PAN-g-PDMAEMA brushes was determined by DLS.     

一种新的岩土流变模型元件

分数阶微积分被发现是一个解决力学建模难题的有力数学工具,利用Riemann-Liouville的分数阶微积分算子及理论,给出一种软体元件及其本构方程,用来模拟介于理想固体和流体之间的土体。该元件能够很好地反映应力松弛和蠕变现象中应力–应变的非线性渐变过程。通过一个软体元件与一个弹簧元件串联或并联,构建2种流变模型,并给出这2种模型的本构方程、松弛模量和蠕变模量。与土的流变实验数据相拟合结果表明,含有软体元件的模型能够更有效地刻画土的流变特性,可以减少参数数量。