Estimation of interregional input–output table using hybrid algorithm of the RAS method and real-coded genetic algorithm

In this paper, we propose a method for improving the accuracy of the estimation of interregional input–output tables, by combining the RAS method and the real-coded Genetic Algorithm (GA); these are simple representative methods for the estimation of an interregional input–output table. By comparing the performance evaluation results obtained using the proposed method, the RAS method, and Simulated Annealing, we verified that the combination of the genetic algorithm and the RAS method can enhance the estimation accuracy of an interregional input–output table. In addition, performance is further enhanced by adjusting GA parameters.     

A Brief Survey of Modern Optimization for Statisticians

Modern computational statistics is turning more and more to high‐dimensional optimization to handle the deluge of big data. Once a model is formulated, its parameters can be estimated by optimization. Because model parsimony is important, models routinely include non‐differentiable penalty terms such as the lasso. This sober reality complicates minimization and maximization. Our broad survey stresses a few important principles in algorithm design. Rather than view these principles in isolation, it is more productive to mix and match them. A few well‐chosen examples illustrate this point. Algorithm derivation is also emphasized, and theory is downplayed, particularly the abstractions of the convex calculus. Thus, our survey should be useful and accessible to a broad audience.     

Neighborhood Effects and Regional Poverty Traps in Rural China

The present paper investigates how neighborhood effects are connected to chronic poverty. We examine a large sample of groups of households and find that neighborhood effects are significant in a majority of groups, especially in the poorest groups. People living in poor communities tend to suffer from poverty over time. It is of theoretical and empirical importance to explore how neighborhood effects are interrelated with chronic poverty and the channels through which this occurs. Unlike other econometric analyses, we establish a multilevel econometric model to show that： （i） it is difficult for an individual living in a neighborhood with a high proportion of agricultural labor, low education levels, and poor transport and telecommunication infrastructure to escape from poverty traps; （ii） neighborhood effects dominate in poor communities; and （iii） although poverty is affected by group-level factors, individual factors still play a dominant role in regards to escaping poverty when income surpasses a threshold level Therefore, policy priority should be given to providing social protection and public services, especially in poor rural areas.     

Vertical Specialization and Trade Growth in Northeast Asia

As the global economy has become further integrated, the international production chain has become more sophisticated, with diversified stages of production located in different countries. Economic theorists have argued that the fragmentation of the global production chain is partly attributable to the high growth in international trade over the past several decades. In this study, we examine vertical specialization in China, Japan and Korea, and its contribution to these nations' trade. Using a multilevel model, it is illustrated that vertical specialization has encouraged increases in trade among all three countries. In particular, China's outcome is remarkable considering how recently it became a member of the WTO.     

提升算法对传统车险索赔频率建模模型的改进——基于我国五省交强险保单数据

广义线性模型作为非寿险定价的经典模型,在非寿险定价中得到了广泛的应用。近年来,以提升算法为代表的机器学习算法在保险领域取得了很好的效果,为保险产品定价提供了一种新的选择。本文将提升算法思想分别融入到回归树模型和广义线性模型(GLM)中去,用得到的新模型对我国车险索赔频率进行预测建模分析,并与传统的回归树模型和GLM进行比较。结果表明,加入提升算法后传统车险索赔频率建模模型的效果得到了很大的改善,并且在不存在过拟合的前提下,随着模型深度和迭代次数的增加,模型的效果也在不断优化。     

使用粒子群算法解决期权定价模型参数校准问题——以heston模型为例

期权定价模型的参数校准问题是一个常见的难题,以heston 模型为例,定价时需要估计6 个参数,参数估计问题实质上是高维非线性规划问题,由于估参函数的性质不好,一般的估参方法常常失效。使用粒子群（PSO）智能算法可以改善该模型的参数校准问题,因为粒子群算法具有内在随机性,因此参数估计中的局部极小值问题可以被较好地解决。使用2017 年12 月20 日的香港恒生指数期权作为估计样本,并对2017 年12 月25 日的期权进行样本外预测,数值结果表明使用heston 模型对期权进行定价并配合粒子群算法估计参数具有良好的定价效果。     

一种基于Edline线特征的车道线识别算法

为了提升车道线检测的准确性和实时性,改良车道偏离预警系统的性能,提出了一种新的车道线识别算法。首先应用投影法对采集到的图像设立感兴趣区域,以此来减少图像中存在的干扰信息;其次应用一种改进后的自适应高斯滤波算法对采集所得图像进行平滑处理,减少图像中不必要的细节;最后采用边缘绘制算法进行边缘检测,在此基础上,提出一种线段检测算法——Edline算法提取边缘线,对检测到的直线段进行筛选和聚类。利用引用计数法对车道线进行跟踪和预测。结果表明,新算法的平均处理时间为17.1 ms,准确率为96.19%,将其应用在车道偏离预警系统中可以有效地提高预警效率,提升预警的准确性和响应速度。研究结果丰富了车道线识别理论,可为车道偏离预警系统的应用及基础研究提供参考。     

On Minimally-supported <Emphasis Type="Italic">D</Emphasis>-optimal Designs for Polynomial Regression with Log-concave Weight Function

This paper studies minimally-supported D-optimal designs for polynomial regression model with logarithmically concave (log-concave) weight functions. Many commonly used weight functions in the design literature are log-concave. For example, and exp(−x ²) in Theorem 2.3.2 of Fedorov (Theory of optimal experiments, 1972) are all log-concave. We show that the determinant of information matrix of minimally-supported design is a log-concave function of ordered support points and the D-optimal design is unique. Therefore, the numerically D-optimal designs can be constructed efficiently by cyclic exchange algorithm.     

机构侵害中小投资者违规监管博弈仿真

本文主要对股票市场中机构投资者侵害中小投资者利益这种违规行为的监管进行了博弈分析,并使用基于主体建模的动态仿真方法在Swarm仿真平台上建立了仿真模型,同时进行了仿真实验。研究指出,严厉的事前监管、完善的上市公司管理制度以及更高的监管者素质可以有效地杜绝机构投资者侵害违规行为的出现。本文通过运用经济仿真方法对股票市场具体违规行为监管问题进行研究,将股票市场监管、博弈论与经济仿真这三个研究领域结合在一起,拓展了对股票市场监管的研究手段,在经济学研究中做出了一定有益的探索。 更多还原     

一种新的卫星通信信号快速盲检测方法

提出了一种卫星通信信号快速盲检测算法,该算法能够适应频谱不重叠的多信号环 境,并能准确估计信号的数量、中心频率及带宽等参数。算法既高效又简单,所需数据量少 且适应低信噪比条件,适于在星载设备上实现。理论分析和仿真及测试结果证明了算法的有 效 性。