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.     

On Quantum Team Games

Recently Liu and Simaan (2004) convex static multi-team classical games have been introduced. Here they are generalized to both nonconvex, dynamic and quantum games. Puu's incomplete information dynamical systems are modified and applied to Cournot team game. The replicator dynamics of the quantum prisoner's dilemma game is also studied.     

不完全信息群体决策专家权重的集结

本文对于属性权重信息和属性效用信息都不完全的群体多属性决策问题，通过构造属性值区间和运用系统聚类分析法，对群体决策中的专家进行分类，并确定每位专家的权重．     

Investigation of Scale Exponents of Complete and Incomplete Aggregation-Annihilation Processes

The complete and incomplete aggregation-annihilation processes are investigated with the method of generating function, and the scale exponents are obtained exactly. We find that the scale exponents of incomplete aggregation-annihilation process are different from the previous exponents obtained by different methods. The time dependence of the total number of clusters and the total mass of clusters are analytically obtained.     

Upper and lower bounds for the energy of bipartite graphs

Using Lagrange's multiplier rule, we find upper and lower bounds of the energy of a bipartite graph G, in terms of the number of vertices, edges and the spectral moment of fourth order. Moreover, the upper bound is attained in a graph G if and only if G is the graph of a symmetric balanced incomplete block design (BIBD). Also, we determine the graphs for which the lower bound is sharp.     

带风险指数的金融均衡的存在性定理及其结构特征(下)

本文主要论证了在不完全市场条件下带风险指数的金融均衡的存在性，并揭示其均衡结构的特征．本文中建立的模型是一、二期货币投入产出金融经济且具有可微的资产结构，这一模型包括了许多具有特殊资产结构的均衡模型，如实资产结构、虚资产结构、恒秩资产结构的均衡模型．因此本文的这一模型具有广泛的应用前景和实用价值．接着给出了本文的金融均衡的存在性定理，再借助微分拓扑给出它的证明过程，这一证明过程较之以前证明均衡存在性的经典方法（如Duffie，D＆W．Shfer（1985）的方法）要简便得多．同时也应注意到本文的这一结论既适用于资产市场下会随机风险因素的情形，也适用于商品空间为无限维的情形，除此之外，还给出了怎样判别资产结构是否属于T类的判别法，为检验均衡存在性提供了更为便利的途径．最后，本文论证了在金融市场里，尽管由于稀缺性的存在，从而导致均衡分配的多样化，然而均衡分配集却形成了一光滑子流，但该流形的维数与稀缺性有关．换句话说，尽管市场是不完全的，但均市分配不确定性的反却是可比的．如此使得人们对均衡资产结构的认识更进一步．     

Evaluation of Orbital- and Ground State Energies of Some Open- and Closed-Shell Atoms over Integer and Noninteger Slater Type Orbitals

Ab initio calculations of the orbital and the ground state energies of some open- and closed-shell atoms over Slater type orbitals with quantum numbers integer and Slater type orbitals with quantum numbers noninteger have been performed. In order to increase the efficiency of these calculations the atomic two-electron integrals were expressed in terms of incomplete beta function. Results were observed to be in good agreement with the literature.     

非完全汽化气相色谱法测定溶剂型涂料中的微量苯、甲苯、二甲苯

用乙酸乙酯稀释溶剂型涂料样品，在较低的汽化温度下直接进样，溶剂型涂料中未汽化的组分留在汽化室里，而苯、甲苯和二甲苯汽化后进入毛细管，用气相色谱法氢火焰离子化检测器进行检测，与国家标准分析方法进行比较，该法具有操作简单、精密度高等优点。     

An open optimization problem in statistical quality control

We ask the experts in global optimization if there is an efficient solution to an optimization problem in acceptance sampling: Here, one often has incomplete prior information about the quality of incoming lots. Given a cost model, a decision rule for the inspection of a lot may then be designed that minimizes the maximum loss compatible with the available information. The resulting minimax problem is sometimes hard to solve, as the loss functions may have several local maxima which vary in an unpredictable way with the parameters of the decision rule.     

Efficiency of connected binary block designs when a single observation is unavailable

In this paper the problem of finding the design efficiency is considered when a single observation is unavailable in a connected binary block design. The explicit expression of efficiency is found for the resulting design when the original design is a balanced incomplete block design or a group divisible, singular or semiregular or regular with ₁>0, design. The efficiency does not depend on the position of the unavailable observation. For a regular group divisible design with ₁>0, the efficiency depends on the position of the unavailable observation. The bounds, both lower and upper, on the efficiency are given in this situation. The efficiencies of designs resulting from a balanced incomplete block design and a group divisible design are in fact high when a single observation is unavailable.The work of the first author is sponsored by the Air Force Office of Scientific Research under Grant AFOSR-90-0092.On leave from Indian Statistical Institute, Calcutta, India. The work of the third author was supported by a grant from the CMDS, Indian Institute of Management, Calcutta.