偏最小二乘回归的贝叶斯正则化神经网络集成模型在证券分析预测中的应用

神经网络集成技术能有效地提高神经网络的预测精度和泛化能力,已经成为机器学习和神经计算领域的一个研究热点.利用Bagging技术和不同的神经网络算法生成集成个体,并用偏最小二乘回归方法从中提取集成因子,再利用贝叶斯正则化神经网络对其集成,以此建立上证指数预测模型.通过上证指数开、收盘价进行实例分析,计算结果表明该方法预测精度高、稳定性好.     

基于BP神经网络的企业未来获利能力智能综合评价

分析了相关分析——多指标综合评价法在确定企业未来获利能力方面的优点和不足 ;并在其基础上提出了基于 BP神经网络的多指标综合评价法 ;仿真试验证明了基于 BP神经网络的多指标综合评价法的有效性     

人工神经网络在SARS疫情分析与预测中的应用

讨论人工神经网络在 SARS疫情分析与预测中的应用 .采用三层结构的反向传播网络 ( Backpropagation network,简称 BP网络 ) ,对 SARS在中国的传播与流行趋势及控制策略建立了网络模型 .并利用实际数据拟合参数 ,针对北京、山西的疫情进行了计算仿真 .结果表明 ,该网络模型算法收敛速度较快 ,预测精度很高     

Asymptotic analysis of a large closed queueing network with discriminatory processor sharing

In this paper the steady-state behavior of a closed queueing network with multiple classes and large populations is investigated. One of the two nodes of the network simply introduces random delays and the discipline in the other node is discriminatory processor sharing. The network is not product-form, so not even the steady-state behavior is known. We assume that the usage is moderately heavy, and obtain two-term asymptotic approximations to the mean number of jobs, and the mean sojourn time, of each class of jobs in the processor node. We also obtain the leading term in the asymptotic approximation to the joint distribution of the number of jobs in the processor node, which is a zero-mean multivariate Gaussian distribution around a line through the origin.     

电力市场中合同电量与竞争电量交易比例的研究

在单边开放的区域电力市场中,合理的合同电量与竞争电量交易比例是保证电力市场有效运行的一个重要环节。竞争电量所占的比例将主要取决于当前发电公司的市场行为。首先使用BP神经网络对电力需求弹性系数进行了预测,然后以长期电力市场均衡为目标函数,考虑贵州电网发电机组的可用容量与负荷预测的误差,以及贵州输电线路的可靠性诸因素,推导出合同电量与竞争电量交易比例,经过与南方区域电力市场目前运营规则规定的交易比例比较,该比例是合理的,可以规避电力市场价格波动等带来的风险。     

Small-world and the growing properties of the Chinese railway network

By using the tools of statistical physics and recent investigations of the scaling properties of different complex networks, the structural and evolving properties of the Chinese railway network (CRN) is studied. It has been verified that the CRN has the same small-world properties of the Indian railway network (IRN). According to the class of small-world networks, we believe the CRN is a single scale. In addition, a novel result is obtained. Measurements on the CRN indicate that the rate at which nodes acquire links depends on the node’s degree and follows a power law.      

综合网管中基于模型驱动的元数据管理

提出了一种基于模型驱动的元数据管理方案．方案采用了基于模型的形式化描述方法，统一了专业网管子系统间的元数据模式；通过引入模型驱动技术，从元模型的抽象层次消除了不同专业网管子系统之间存在的“信息孤岛”；方案还对综合网管环境下的元数据理论和交换机制进行了研究，并给出了元数据的管理框架以及元数据交换的Java接口映射．     

Territories,corridors, and networks: A biological model for the premodern state

When depicted on maps as homogenous territorial wholes, ancient states are visually summarized as static entities in a way that conceals the highly fluid dynamics of polity formation, maintenance, and growth. Models derived from studies of animal behavior show that “territory” does not consist of an undifferentiated use of the landscape. Instead, the concept of territory can be parsed into a series of resource‐rich nodes linked by corridors of access, surrounded by unutilized regions and boundaries marked at points of competition. Ancient human groups also can be analyzed as having perceived and occupied landscapes through strategies of flexible networks in which nodes and corridors were surrounded by unutilized spaces around which boundaries were selectively identified and defended. This strategy is identifiable in human social groups at different levels of complexity ranging from hunter‐gatherers through ancient chiefdoms and states. © 2007 Wiley Periodicals, Inc. Complexity 12: 28–35, 2007     

A novel approach based on recurrent neural networks applied to nonlinear systems optimization

This paper presents an efficient approach based on recurrent neural network for solving nonlinear optimization. More specifically, a modified Hopfield network is developed and its internal parameters are computed using the valid subspace technique. These parameters guarantee the convergence of the network to the equilibrium points that represent an optimal feasible solution. The main advantage of the developed network is that it treats optimization and constraint terms in different stages with no interference with each other. Moreover, the proposed approach does not require specification of penalty and weighting parameters for its initialization. A study of the modified Hopfield model is also developed to analyze its stability and convergence. Simulation results are provided to demonstrate the performance of the proposed neural network.