Multi-label dimensionality reduction and classification with extreme learning machines

In the need of some real applications, such as text categorization and image classification, the multi-label learning gradually becomes a hot research point in recent years. Much attention has been paid to the research of multi-label classification algorithms. Considering the fact that the high dimensionality of the multi-label datasets may cause the curse of dimensionality and wil hamper the classification process, a dimensionality reduction algorithm, named multi-label kernel discriminant analysis （MLKDA）, is proposed to reduce the dimensionality of multi-label datasets. MLKDA, with the kernel trick, processes the multi-label integrally and realizes the nonlinear dimensionality reduction with the idea similar with linear discriminant analysis （LDA）. In the classification process of multi-label data, the extreme learning machine （ELM） is an efficient algorithm in the premise of good accuracy. MLKDA, combined with ELM, shows a good performance in multi-label learning experiments with several datasets. The experiments on both static data and data stream show that MLKDA outperforms multi-label dimensionality reduction via dependence maximization （MDDM） and multi-label linear discriminant analysis （MLDA） in cases of balanced datasets and stronger correlation between tags, and ELM is also a good choice for multi-label classification.     

横坐标系捷联惯导系统极区导航及阻尼设计

针对水下潜器在极区工作时,其常用惯性导航系统机械编排存在精度下降、无北向基准等问题,设计了适用于水下潜器极区工作的横坐标系捷联惯性导航系统机械编排方案。在构建了横坐标系参考框架的基础上,推导出各导航参数在横坐标系统和常规坐标系统之间的转换关系,建立了横坐标系捷联惯导系统的机械编排,并在此基础上分析其误差传播特性。通过误差分析,确定了横坐标系捷联惯导系统中存在的周期性振荡,并提出了适用于横坐标系捷联惯导系统的阻尼技术,有效抑制了周期性振荡。最后,通过仿真试验验证了该系统在极区工作的可行性和阻尼技术有效性。     

Evidence for self-organized criticality in unscheduled downtime data of equipment

How to reduce downtime and improve availability of the complex equipment is very important. Although the unscheduled downtime（USDT） issues of the equipment are very complex, the self-organized criticality（SOC） is the right theory to study complex systems evolution and opens up a new window to the investigation of disasters, such as the sudden failure of the equipment. Firstly,SOC theory and its validation method are introduced. Then an SOC validation method for USDT of the equipment is proposed based on the above theory. Case study is done on bottleneck equipment in a factory and corresponding data pre-process work is done. The rescaled-range（R/S） analysis method is used to calculate the Hurst exponent of USDT time-series data in order to determine the long-range correlation of USDT data on time scale;at the same time the spatial power-law characteristic of USDT time series data is studied. The result shows that the characteristics of SOC are revealed in USDT data of the equipment according to the criterion of SOC. In addition, based on the characteristics of SOC,the overall framework of the prediction method for major sudden failure of the equipment is proposed based on SOC.     

Information identification in different networks with heterogeneous information sources

Traditional cheaptalk game model with homogeneous information sources provided a con clusion that dishonest information sources will not be identified if he changes strategy stochastically. In this paper, the authors incorporate different information diffusion networks and heterogeneous in formation sources into an agentbased artificial stock market. The obtained results are different with traditional results that identification ability of uninformed agents has been highly improved with diffu sion networks and heterogeneous information sources. Additionally, the authors find uninformed agents can improve identification ability only if there exists a sufficient number of heterogeneous information sources in stock market.     

求解双层CARP优化问题的知识型遗传算法

双层CARP优化问题首先考虑物流系统的宏观配置问题,然后考虑相关服务的完成问题. 提出了一种求解双层CARP优化问题的知识型遗传算法：采用两种扩展启发式方法来辅助生成初始种群;基于算子绩效知识为每次选择、交叉和变异操作选择合适的操作算子;基于弧段顺序知识为每次交叉和变异操作选择合适的断点位置;采用局部替换程序不断向当前种群中注入新个体. 实验结果表明,知识型遗传算法在优化性能方面优于其他几种方法.     

软件可靠性关键参数判定的矩估计方法

已有的基于结构的软件可靠性研究中大多假设建模参数为已知常量,导致模型忽略了参数估计的统计特征对于系统可靠性的影响.针对该问题,提出一种基于矩估计的软件可靠性关键参数判定方法,该方法通过分析基于结构软件可靠性模型的特点,建立各建模参数的矩估计量与系统可靠性预测结果之间的关系,量化各参数对于系统可靠性的影响,判定关键参数.为系统可靠性控制、系统结构优化及软件测试资源最优化分配等问题的解决提供了理论支持,通过对两个实际软件系统的分析验证了方法的有效性.     

Empirical likelihood inference for logistic equation with random perturbation

Empirical likelihood (EL) combined with estimating equations (EE) provides a modern semi-parametric alternative to classical estimation techniques such as maximum likelihood estimation (MLE). This paper not only uses closed form of conditional expectation and conditional variance of Logistic equation with random perturbation to perform maximum empirical likelihood estimation (MELE) for the model parameters, but also proposes an empirical likelihood ratio statistic (ELRS) for hypotheses concerning the interesting parameter. Monte Carlo simulation results show that MELE and ELRS provide competitive performance to parametric alternatives.     

基于变分推断的一般噪声自适应卡尔曼滤波

线性高斯状态空间模型中假设噪声为已知的白噪声过于苛刻。认为过程噪声与观测噪声均未知且二者的解析关系确定,假设观测噪声的均值非零且服从高斯分布,方差服从逆威沙特分布,从而构成了层次式贝叶斯模型。利用变分推断将均值与方差和系统状态一起作为随机变量进行迭代估计,在得到观测噪声的均值与方差的估计值后,利用其与过程噪声的关系进一步更新未知过程噪声的均值与方差,从而动态地得到每一时刻过程噪声与观测噪声的一、二阶统计矩信息,即使在噪声统计信息动态变化的情况下也有较满意的滤波精度。实验证明了该算法的有效性。     

Calculation of Turbulent Flow with Pseudo-spectral Method

This paper deals with the numerical simulation of incompressible turbulent boundary flow of a flat plate with the pseudo-spectral matrix method. In order to appear more than 10 nodes in the turbulent base-stratum and transition of 43×43 computational grids,a coordinate transformation is put up from physical panel to computational panel. Several zero turbulent models are computed comparatively. The results are credible when comparing with the previous methods.     

位于强震区的农村中小学教学楼砌体结构抗震设计分析

针对大开间少墙砌体结构的受力特点，结合教学楼震害，分析了构造柱、圈梁对砌体结构抗震性能的影响，并对提高结构延性的设计理念进行了阐述；同时对大开间少墙单面走廊砌体结构的设计提出了几点建议。