排名聚合算法在少量长列表聚合中的性能比较分析

排名聚合将多个排名列表聚合成一个综合排名列表,可应用于推荐系统、链路预测、元搜索、提案评选等.当前已有工作从不同角度对不同排名聚合算法进行了综述、比较,但存在算法种类较少、数据统计特性不清晰、评价指标不够合理等局限性.不同排名聚合算法在提出时均声称优于已有算法,但是用于比较的方法不同,测试的数据不同,应用的场景不同,因此何种算法最能适应某一任务在很多情况下仍不甚清楚.本文基于Mallows模型,提出一套生成统计特性可控的不同类型的排名列表的算法,使用一个可应用于不同类型排名列表的通用评价指标,介绍9种排名聚合算法以及它们在聚合少量长列表时的表现.结果发现启发式方法虽然简单,但是在排名列表相似度较高、列表相对简单的情况下,能够接近甚至超过一些优化类方法的结果;列表中平局数量的增长会降低聚合排名的一致性并增加波动;列表数量的增加对聚合效果的影响呈现非单调性.整体而言,基于距离优化的分支定界方法 (FAST)优于其他各类算法,在不同类型的排名列表中表现非常稳定,能够很好地完成少量长列表的排名聚合.     

Gaussian states as minimum uncertainty states

In bosonic fields, Gaussian states, which consist of a rather wide family of states including coherent states, squeezed states, thermal states, etc., have many classical-like features, and are usually defined from the mathematical perspective in terms of characteristic functions. It is well known that some special Gaussian states, such as coherent states, are minimum uncertainty states for the conventional Heisenberg uncertainty relation involving canonical pair of position and momentum observables. A natural question arises as whether all Gaussian states can be characterized as minimum uncertainty states. In this work, we show that indeed Gaussian states coincide with minimum uncertainty states for an information-theoretic refinement of the conventional uncertainty relation established in Luo (2005) [40]. This characterization puts Gaussian states on a novel basis of physical significance.     

拉普拉斯约束低秩表示的高光谱图像异常检测

伴随高光谱图像的广泛使用,高光谱图像技术得到长足的发展,其中高光谱图像异常检测技术越发受到重视。为了解决传统高光谱图像异常检测技术的实用性和检测效果不佳的问题,提出一种新颖的低秩表示检测算法。对于高光谱图像,大部分背景像元均可以被少量主要的背景像元组合近似地表示,且它们的表示系数将会位于低秩的空间中。在剩下无法被主要背景像元表示的稀疏部分中存在着异常像元,则可以被检测算法提取出来。在低秩表示中,背景像元字典的构建将会影响高光谱图像中背景像元的表示。如直接从现有高光谱图像中提取背景像元构建字典,会导致异常像元对背景像元字典的污染。而利用待检测高光谱图像观测数据和由光谱组成原理可合成的潜在未观测数据来构建背景像元字典,提取出背景像元的主要特征,有利于更好地分离出稀疏异常像元的信息。并且高光谱图像数据存在高维几何结构特点,通过引入拉普拉斯矩阵来约束空间中局部相似的像元对于待检测像元的表示作用,获得更接近于真实的表示系数。实验结果分别在仿真数据和真实数据上验证,与传统方法相比,提出的方法通过有效地突出异常像元提高了检出率和抑制了背景像元,降低了误检率。     

Rank-based deactivation model for networks with age

We study the impact of age on network evolution which couples addition of new nodes and deactivation of old ones. During evolution, each node experiences two stages: active and inactive. The transition from the active state to the inactive one is based on the rank of the node. In this paper, we adopt age as a criterion of ranking, and propose two deactivation models that generalize previous research. In model A, the older active node possesses the higher rank, whereas in model B, the younger active node takes the higher rank. We make a comparative study between the two models through the node-degree distribution.     

Studies on controllability of directed networks with extremal optimization

Almost all natural, social and man-made-engineered systems can be represented by a complex network to describe their dynamic behaviors. To make a real-world complex network controllable with its desired topology, the study on network controllability has been one of the most critical and attractive subjects for both network and control communities. In this paper, based on a given directed–weighted network with both state and control nodes, a novel optimization tool with extremal dynamics to generate an optimal network topology with minimum control nodes and complete controllability under Kalman’s rank condition has been developed. The experimental results on a number of popular benchmark networks show the proposed tool is effective to identify the minimum control nodes which are sufficient to guide the whole network’s dynamics and provide the evolution of network topology during the optimization process. We also find the conclusion: “the sparse networks need more control nodes than the dense, and the homogeneous networks need fewer control nodes compared to the heterogeneous” (Liu et al., 2011  [18]), is also applicable to network complete controllability. These findings help us to understand the network dynamics and make a real-world network under the desired control. Moreover, compared with the relevant research results on structural controllability with minimum driver nodes, the proposed solution methodology may also be applied to other constrained network optimization problems beyond complete controllability with minimum control nodes.     

Entropy in linear programs

This paper treats entropy constrained linear programs from modelling as well as computational aspects. The optimal solutions to linear programs with one additional entropy constraint are expressed in terms of Lagrange-multipliers. Conditions for uniqueness are given. Sensitivity and duality are studied. The Newton—Kantorovich method is used to obtain a locally convergent iterative procedure. Related problems based on maximum entropy or minimum information are discussed.     

最小方差无失真响应浊音谱建模方法研究

讨论了在语音处理中广泛使用的线性预测模型,分析了线性预测模型在建模语音谱包络方面存在缺点,并分析了线性预测模型过高估计浊音谐波频率处能量以及随着阶数增加线性预测谱逐渐恶化的原因。在此基础上,提出了最小方差无失真响应(MVDR)建模方法,并讨论了通过Toeplitz矩阵的Cholesky分解确定MVDR滤波器系数的方法,通过与线性预测方法进行比较,发现最小方差无失真响应滤波器能提供一个更好的原始语音包络。     

Antifungal activity of multifunctional Fe3O4-Ag nanocolloids

In recent years, rapid increase has been observed in the population of microbes that are resistant to conventionally used antibiotics. Antifungal drug therapy is no exception and now resistance to many of the antifungal agents in use has emerged. Therefore, there is an inevitable and urgent medical need for antibiotics with novel antimicrobial mechanisms. Aspergillus glaucus is the potential cause of fatal brain infections and hypersensitivity pneumonitis in immunocompromised patients and leads to death despite aggressive multidrug antifungal therapy. In the present article, we describe the antifungal activity of multifunctional core-shell Fe₃O₄-Ag nanocolloids against A. glaucus isolates. Controlled experiments are also carried out with Ag nanocolloids in order to understand the role of core (Fe₃O₄) in the antifungal action. The minimum inhibitory concentration (MIC) of nanocolloids is determined by the micro-dilution method. MIC of A. glaucus is 2000 μg/mL. The result is quite promising and requires further investigations in order to develop a treatment methodology against this death causing fungus in immunocompromised patients.     

铁基超导体Minimum模型中的d波和扩展s波

对比研究了Minimum模型中d-波配对和扩展s-波配对的基本性质.结果表明,超导序参量按类BCS关系随温度变化,重整化参数t对超导有抑制作用,当t足够大时,超导消失,系统进入正常金属相.此外,讨论了超导序参量随不同参量的变化关系.对于d波,系统有稳定态;而对于扩展s波,系统则没有稳定态.