Integrating marketing databases through regressed microsegmentation

In order to quantitatively target media to markets, media planners need a data base that contains both target market membership and media usage variables. In the absence of single-source data, containing both kinds of variables, researchers have sought to synthesize, or integrate, these data through a variety of different approaches. One of the most elegant is microsegmentation. While the approach has been shown to consistently underestimate the concentration of target market members in the most selective media vehicles, this study suggests that the method might produce a proper ordering, or better yet, a proportionate evaluation of media. This, in turn, suggests that a regression equation might be used to adjust initial estimates of target market concentrations, thus providing a better means of linking target market and media data bases.     

基于快乐指数的“个人–团队”动态匹配模型研究——以科研团队为例

个人–团队匹配模型已经改变了团队模式下人力资源的选拔方法,但不足的是不能保证"个人–团队"能够动态匹配。已有的研究发现,员工快乐与否直接影响着团队的绩效,并且员工快乐这个元素是呈动态变化的。建立了团队快乐指数的评价指标体系,并通过构建"个人–团队"动态匹配模型来影响团队成员快乐的因素,达到间接、动态地调整"个人–团队"绩效关系,使个人与团队之间得到双赢的绩效目标。     

Project Matching and Overcapacity in China

Based on the features of China’s project investment, we consider the formation of production capacity as a matching behavior between local governments and investment enterprises. Using the search and matching model, we illustrate that the excess capacity in China mainly results from the asymmetry between the gains from and contribution to the project matching: The capacity will be excessive when the proportion of local governments’ return exceeds its contribution to the project, and the more unbalanced the return–contribution relationship, the more severe the overcapacity. Meanwhile, we test this theoretical prediction based on a quasi-natural experiment: the reform of administrative approval system. The empirical results show that the reduction of the local governments’ return–contribution ratio will significantly raise the capacity utilization rate and mitigate the overcapacity. Industry-specific regression results further indicate that governments’ return–contribution asymmetry is more prominent in industries dominated by state-owned enterprises, high-monopoly industries, heavy industries, and industries with serious overcapacity. This paper offers a novel mechanism of overcapacity, a theoretical criterion for judging optimal capacity, and some new regulatory tools with the micro foundation.     

Do mobile banner ads increase sales? Yes,in the offline channel

Firms allocate increasingly large budgets to mobile banner advertising. Yet, existing research paid only scant attention to the sales effects of mobile banner ads. In this paper, we fill this gap by determining the offline and online sales impact of a large-scale mobile banner advertising campaign. As part of a geographical field experiment, over 3.5 million mobile banner ads were served to a predetermined geographical area. We determine the offline and online sales effects of the mobile banner ad campaign by analyzing twenty months of sales data for regions covering the entire country of the Netherlands. Relying on a difference-in-difference approach and two matching methods, we demonstrate an offline sales increase of around 2%. The online sales effect is not significant. We conclude that firms can use mobile banner advertising to boost offline sales. We find no evidence for cross-channel sales cannibalization.     

多视角穿墙雷达联合稀疏成像算法

多视角穿墙雷达成像系统利用多个视角的目标回波数据,可以有效提高目标成像重建结果的质量。在建立多视角穿墙雷达联合稀疏信号模型的基础上,提出了一种基于交叉验证技术的删失同时正交匹配追踪成像算法。该算法将每个观测视角雷达单元的测量数据分成重建数据和交叉验证数据两部分,通过进行多次删失同时正交匹配追踪迭代计算实现测量噪声水平估计和成像重建,既减小了各个视角雷达站间的数据通信开销,也摆脱了成像算法对测量噪声水平和场景稀疏度先验信息的依赖。仿真实验结果验证了所提成像算法的有效性和准确性。     

基于压缩感知的伪随机等效采样信号重构

伪随机等效采样利用采样周期数与采样点数间的互质关系使各采样点均匀复现于同一周期,从而达到较高的等效采样速率。然而为了精确重构出原始信号,需大量采样数据,因此导致采样时间过长,实时性能差。针对上述问题,提出了一种基于压缩感知理论的伪随机等效采样信号重构方法,通过构造伪随机等效采样观测矩阵并选择离散傅里叶变换基建立稀疏重构模型,然后利用压缩感知中的正交匹配追踪算法求解该模型,从而重构出原始信号。仿真实验表明,所提方法在采样点个数40时,重构成功率达99.73%。     

创业决策机理研究：影响因素与作用方式*

本文以情境检测的方式探讨创业决策的影响因素与作用机理,通过对364份样本的分析,发现已有研究所关注的风险感知、风险倾向、手段导向逻辑、经验认知等因素,都对创业决策具有重要影响。与已有研究结论不同的是,本文发现风险感知不是风险倾向影响创业决策的中介变量,而是与风险倾向并列影响创业决策;与二者并列的另一个重要因素,是近年来创业研究所关注的手段导向逻辑。研究结果发现,经验认知作为自变量,通过风险感知、风险倾向、手段导向逻辑对创业决策发挥影响。     

Evaluating the effects of equity incentives using PSM: Evidence from China

This paper investigates the effects of equity incentives on firm performance in Chinese listed firms. We address the sample selection problem by employing the propensity score matching methodology. Results show that, (1) On the whole, performance is positively related to equity incentives even after controlling for sample selection bias; (2) The final control rights have an important impact on the effects of equity incentives. The execution of equity incentives in privately owned firms can significantly decrease the agency costs between shareholders and managers, but such effects cannot be observed in state-owned firms; (3) Effects of equity incentives depend on the incentive type, that is, comparing to stock-based incentives, option-based incentives can reduce the agency costs significantly, thus are more effective; (4) Ownership structure also has important impacts on the effects of equity incentives. The agency costs decrease in firms with more decentralized ownership after introducing equity incentive, while in concentrated firms the effect is negligible.     

网络经济下中小企业理论探索

中国股市的财富效应很不显著。因此,期望通过“激活股市刺激消费”的政策主张目前难以奏效。笔者认为,要尽快培育和发挥股市刺激消费与促进经济增长的作用,必须把通过股市“富民”的方略置于重要位置。     

基于权变融资需求的VC-E云计算匹配方法

近年来,风险投资行业聚焦于投资者与风险公司之间的匹配问题.但学者们的目光一直致力于使用罗斯的双边匹配方法来研究风险公司与投资者之间的双边匹配.随着金融技术发展及进步,以及算法经济时代的来临,由于金融科技的影响,传统金融业已经发生了根本性的变化.紧跟时代的变化,利用信息服务的动态匹配模型研究云计算影响下风险投资者与创业公司的匹配方法,企图达到两个目的:改变创业企业多轮融资现状,使其可以实现无阶段的随需随融;改变区域性的融资现状,使融资可以实现互联网下的全球无界融资.在新业态层面上,这对风险性的企业权变融资具有重要参考意义.