光盘用光学头物镜性能的综合评估

论述了采用波象差均方根、点扩展函数、光学传递函数对高次旋转对称非球面光学头物镜性能进行综合评估的方法。作者在正则坐标理论基础〔１〕上编制了光学头物镜性能综合评估软件。给出了评估实例。     

On the monotonicity of (LDL) logic programs with set

LDL is one of the recently proposed logical query languages, which incorporate set, for data and knowledge base systems. Since LDL programs can simulate negation, they are not monotonic in general. On the other hand, there are monotonic LDL programs. This paper addresses the natural question of “When are the generally nonmonotonic LDL programs monotonic?” and investigates related topics such as useful applications for monotonicity. We discuss four kinds of monotonicity, and examine two of them in depth. The first of the two, called “ω-monotonicity”, is shown to be undecidable even when limited to single-stratum programs. The second, called “uniform monotonicity”, is shown to implyω-monotonicity. We characterize the uniform monotonicity of a program (i) by a relationship between its Bancilhon-Khoshafian semantics and its LDL semantics, and (ii) with a useful property called subset completion independence. Characterization (ii) implies that uniformly monotonie programs can be evaluated more efficiently by discarding dominated facts. Finally, we provide some necessary and/or sufficient, syntactic conditions for uniform monotonicity. The conditions pinpoint (a) enumerated set terms, (b) negations of membership and inclusion, and (c) sharing of set terms as the main source for nonuniform monotonicity.     

夜视系统鉴别率测试的新方法

用室内低光照高对比鉴别率取代野外探测目标距离作为衡量夜视系统性能优劣的标准,将给验收工作带来方便。本文给出一种实验室内测试夜视系统鉴别率的方法。文中导出了夜视系统鉴别率的一种新的表达式,从而提出了一种改进后的物理模型及其实验方案。分析表明,采用新方案,即在实验室中人为增加环境照度,由此测得的系统工作鉴别率,用以取代野外探测目标距离,其可信度可望得到明显的提高。     

A hybrid algorithm for Bayesian network structure learning with application to multi-label learning

We present a novel hybrid algorithm for Bayesian network structure learning, called H2PC. It first reconstructs the skeleton of a Bayesian network and then performs a Bayesian-scoring greedy hill-climbing search to orient the edges. The algorithm is based on divide-and-conquer constraint-based subroutines to learn the local structure around a target variable. We conduct two series of experimental comparisons of H2PC against Max–Min Hill-Climbing (MMHC), which is currently the most powerful state-of-the-art algorithm for Bayesian network structure learning. First, we use eight well-known Bayesian network benchmarks with various data sizes to assess the quality of the learned structure returned by the algorithms. Our extensive experiments show that H2PC outperforms MMHC in terms of goodness of fit to new data and quality of the network structure with respect to the true dependence structure of the data. Second, we investigate H2PC’s ability to solve the multi-label learning problem. We provide theoretical results to characterize and identify graphically the so-called minimal label powersets that appear as irreducible factors in the joint distribution under the faithfulness condition. The multi-label learning problem is then decomposed into a series of multi-class classification problems, where each multi-class variable encodes a label powerset. H2PC is shown to compare favorably to MMHC in terms of global classification accuracy over ten multi-label data sets covering different application domains. Overall, our experiments support the conclusions that local structural learning with H2PC in the form of local neighborhood induction is a theoretically well-motivated and empirically effective learning framework that is well suited to multi-label learning. The source code (in R) of H2PC as well as all data sets used for the empirical tests are publicly available.     

一种关于目标罚参数的精确罚函数法

罚函数法是一种将约束优化问题转化为无约束问题的重要方法.对于一般的约束优化问题,通过加入新参数,给出了一种改进的精确罚函数和这种罚函数的精确罚定理证明,提出了求解这种罚函数的算法.实验表明该算法是有效的.     

基于BP神经网络的手写数字识别的算法

由于BP神经网络具有并行处理信息、自组织、自学习信息等优点,本文采用了BP神经网络对手写数字识别进行运算,提取笔画密度、长宽比和欧拉数等特征作为训练样本.并用Matlab对其算法进行仿真,并且很准确的识别出来,说明其有非常广泛的前景.     

加强企业内部经济责任制运行之我见

企业内部经济责任制是众多企业经营管理工作的一项基本管理制度。指标合理、体系完备、主体明确是推行责任制的基础;强化过程监控是保障经济责任制有效运行的必要手段;强化目标管理、严格考核兑现是推行经济责任制的关键。只有科学确定企业总体经营目标、合理分解目标任务、考核奖励公平公正,才能保证企业经济责任制的顺利实施。     

环状扫描的强深度森林

深度森林（Deep Forest,DF）,由于此模型超参数少,且参数设置没有过多的要求,训练方便,鲁棒性高,因此在处理大型数据时比神经网络算法更加具有优势。但是,传统的深度森林中,多粒度扫描忽略了边缘数据携带的隐含信息,无法充分地获得各个特征子集,进而会对以后的级联部分产生影响。而且,级联部分每次得到的新特征有限,影响了模型的表征学习能力。针对以上问题,提出一种环状强深度森林（Circular Strong Deep Forest,CSDF）,其通过环状扫描过程,一定程度上得到更充分的特征子集,且强级联森林通过特征选择提高了模型的表征学习能力。经过在不同数据集上的测试,结果表明,CSDF的性能更加优越,尤其是高维数据上更为明显。     

Security optimal power flow considering loading margin stability using hybrid FFA–PS assisted with brainstorming rules

This paper presents a new power system planning strategy which combines firefly algorithm (FFA) with pattern search algorithm (PS). The purpose is minimizing total fuel cost, total power loss and reducing total voltage deviation, with the objective of enhancing the loading margin stability and consequently the power system security. A new interactive and simple mechanism, inspired in brainstorming process, is proposed that allows FFA and PS algorithms to explore new regions of the search space. In this study the Static VAR compensator (SVC) is modeled and integrated in an efficient location which is chosen considering the voltage stability index. The proposed algorithm is interactive and tries to optimize a set of control variables at the same time, namely, active power generations, voltage of generators, tap transformers, and the reactive power of shunt compensators to optimize three objective functions such as: fuel cost, total power loss and total voltage deviation. These variables are optimized using a flexible interactive and competitive search mechanism. The proposed planning strategy has been examined and applied to two practical test systems IEEE 14-Bus and IEEE 30-Bus. Simulation results confirm the effectiveness of this hybrid strategy for solving the security optimal power flow.