求半正定Hermitian矩阵特征向量的改进乘幂法

经典乘幂法结合压缩法是计算半正定Hermitian矩阵最大前n个特征值对应特征向量的重要方法,但其固定的迭代次数使得待分解矩阵的随机变化和初始向量的不同选择导致计算精度波动较大,同时,较大特征值对应特征向量的计算误差也会影响较小特征值对应特征向量的计算。为克服这些缺点,提出了一种将前后两次迭代所求向量的距离作为迭代终止条件的改进乘幂法,并证明了它在有误差传播时的收敛性。理论计算结果表明,对4阶半正定Hermitian随机矩阵,在相同计算精度前提下,所提方法比经典方法可至少降低一半计算复杂度。     

异质性贸易成本与企业出口产品质量:机理与事实

文章基于Antoniades（2015）模型,将企业创新行为纳入企业产品质量模型,分析贸易成本变化对企业出口产品质量影响的微观机理,发现贸易成本上升对企业出口产品质量同时存在生产率门槛效应和创新抑制效应,生产率门槛效应会放大在位企业的生产率优势,促进企业出口产品质量提升;而创新抑制效应会削弱企业的创新水平,降低企业出口产品质量。采用中国微观企业数据检验相关结论发现：总体上,贸易成本上升显著的降低了企业出口产品质量,其中,贸易成本对劳动密集型企业和加工贸易企业的出口产品质量降低作用更大,而对资本密集型企业和一般贸易企业的降低幅度较小。中介效应的检验发现：贸易成本的生产率门槛效应和创新抑制效应同时存在,并且创新抑制效应更大,因此总体上,贸易成本会降低企业出口产品质量。     

Modelling the housing market in OECD countries

Recent episodes of housing bubbles, which occurred in several economies after the burst of the United States housing market, suggest studying the evolution of housing prices from a global perspective. We utilise a theoretical model for the purposes of this contribution, which identifies the main drivers of housing price appreciation, such as, for example, income, residential investment, financial elements, fiscal policy and demographics. In a second stage of our analysis, we test our theoretical hypothesis by means of a sample of 18 OECD countries from 1970 to 2011. We employ the vector error correction econometric technique in terms of our empirical analysis, which permits us to model the long-run equilibrium relationship and the short-run dynamics, which also helps to account for endogeneity and reverse causality problems.     

一种最大分类间隔SVDD的多类文本分类算法

文本分类是信息检索和文本挖掘的关键技术之一。提出了一种基于支持向量数据描述(SVDD)的多类文本分类算法,用支持向量描述训练求得包围各类样本的最小超球体,并使得分类间隔最大化,在测试阶段,引入基于核空间k-近邻平均距离的判别准则,判断样本所属类别。实验结果表明,该方法具有很好的泛化能力和很好的时间性能。     

AdaBoost集成SVM的供应链金融信用风险评估

为了实现供应链金融信用风险的科学定量管理,建立了一种AdaBoost集成的支持向量机（SVM）算法。该方法首先采用SVM方法对信用风险数据进行分类学习,建立基学习器;接着通过AdaBoost集成算法对基学习器迭代训练,生成最终的供应链金融信用风险评估模型。实证结果表明,AdaBoost集成SVM分类器较模糊积分SVM集成等方法具有更高的分类准确率,因此该模型具有很好的应用前景。     

基于相空间重构及自适应支持向量机的短期风速预测

风速具有较强的随机性和间歇性,导致大规模风电接入电网会严重影响电力系统的安全稳定运行以及电能质量。较为准确的风速预测可以降低风能对电网的不利影响,为电网运行调度提供可靠的依据。在对风速进行混沌属性分析及相空间重构的基础上,采用自适应支持向量机进行短期风速预测,结果表明该方法的预测精度高于BP、RBF等预测模型。     

House Prices and Stock Prices: Different Roles in the US Monetary Transmission Mechanism*

We analyze the role of house prices and stock prices in the monetary‐policy transmission mechanism in the US, using a structural vector autoregressive model. If we allow the interest rate and asset prices to react simultaneously to news, we find different roles for house prices and stock prices in the monetary transmission mechanism. Following a contractionary monetary‐policy shock, stock prices fall immediately, while the response in house prices is more gradual. Regarding the systematic response in monetary policy, stock prices play a more important role than house prices. As a consequence, house prices contribute more than stock prices to fluctuations in gross domestic product and inflation.     

Experimental evidence on the ‘insidious’ illiquidity risk

This paper introduces an experiment aiming to investigate the contribution of illiquidity risk to the total risk of a collective investment project. If implemented, the project succeeds with a known probability. Yet the project fails if the quota of investors is not reached in the first place. Hence strategic uncertainty compounds its effect with the “intrinsic risk” of the project. Results confirm the insidious nature of illiquidity: as long as a first collective default does not occur, investors accept high intrinsic risk projects. After a first default, they become extremely prudent and come back to market only gradually. After several defaults, private agents manage to coordinate on a relatively low intrinsic risk above which they refuse to participate in the project. Macroeconomic policy implications follow.     

Debt, growth and inflation in large European economies: a vector auto-regression analysis

This paper attempts to empirically test the hypothesis that whether debt matters in the EU. This has been performed by examining the potential adverse effects of debt in large European economies on investment, inflation and growth. Using the hybrid cointegration and vector autoregressive models, the findings, based on the period 1970–97, suggest that debt causes significant adverse effects on investment, but its impact on growth is not clear-cut. Moreover, debt appears to be inflationary in most cases in the long run, though produces no clear short run pattern on inflation.     

数字线化图（DLG）要素分类探讨

探讨数字线划图的特点以及编辑要点。