Nowcasting South African gross domestic product using a suite of statistical models

Given lags in the release of data, a central bank must ‘nowcast’ current gross domestic product (GDP) using available quarterly or higher frequency data to understand the current state of economic activity. This paper uses various statistical modelling techniques to draw on a large number of series to nowcast South African GDP. We also show that GDP volatility has increased markedly over the last 5 years, making GDP forecasting more difficult. We show that all the models developed, as well as the Reserve Bank's official forecasts, have tended to overestimate GDP growth over this period. However, several of the statistical nowcasting models we present in this paper provide competitive nowcasts relative to the official Reserve Bank and market analysts' nowcasts.     

赋能型领导对员工创造力的影响：个人学习与成就动机的作用

以创造力理论为基础,探究赋能型领导对员工创造力的作用机制,检验个人学习的中介作用和成就动机的调节作用。通过采用验证性因子分析、层次回归分析和Bootstrap法,对收集的309份问卷进行分析。研究结果表明:①赋能型领导对员工创造力有正向影响作用;②个人学习在赋能型领导和员工创造力之间发挥部分中介作用;③成就动机调节了赋能型领导与员工个人学习之间的关系,且被调节的中介模型得到了验证。     

Incorporating textual information in customer churn prediction models based on a convolutional neural network

This study investigates the value added by incorporating textual data into customer churn prediction (CCP) models. It extends the previous literature by benchmarking convolutional neural networks (CNNs) against current best practices for analyzing textual data in CCP, and, using real life data from a European financial services provider, validates a framework that explains how textual data can be incorporated in a predictive model. First, the results confirm previous research showing that the inclusion of textual data in a CCP model improves its predictive performance. Second, CNNs outperform current best practices for text mining in CCP. Third, textual data are an important source of data for CCP, but unstructured textual data alone cannot create churn prediction models that are competitive with models that use traditional structured data. A calculation of the additional profit obtained from a customer retention campaign through the inclusion of textual information can be used by practitioners directly to help them make more informed decisions on whether to invest in text mining.     

The role of text-extracted investor sentiment in Chinese stock price prediction with the enhancement of deep learning

Whether investor sentiment affects stock prices is an issue of long-standing interest for economists. We conduct a comprehensive study of the predictability of investor sentiment, which is measured directly by extracting expectations from online user-generated content (UGC) on the stock message board of Eastmoney.com in the Chinese stock market. We consider the influential factors in prediction, including the selections of different text classification algorithms, price forecasting models, time horizons, and information update schemes. Using comparisons of the long short-term memory (LSTM) model, logistic regression, support vector machine, and Naïve Bayes model, the results show that daily investor sentiment contains predictive information only for open prices, while the hourly sentiment has two hours of leading predictability for closing prices. Investors do update their expectations during trading hours. Moreover, our results reveal that advanced models, such as LSTM, can provide more predictive power with investor sentiment only if the inputs of a model contain predictive information.     

Managers’ exploration activities and individual unlearning: the mediating role of learning orientation and reflection

Abstract

The main goal of this study was to investigate the effects of managers’ exploration activities on learning orientation, reflection, and unlearning of team members. A questionnaire survey was conducted among 115 employees in 23 teams from a Japanese pharmaceutical company. The results of multi-level analyses suggest that managers’ exploration activities had a direct influence on team members’ learning orientations, which subsequently had a positive influence on their unlearning, with and without the mediation of reflection. These findings suggest that a subordinate’s unlearning is driven by managers’ activities through motivational and cognitive processes. The theoretical and practical implications of the findings are discussed.     

Resilience of sustainability‐oriented and financially‐driven organizations

To answer key questions concerning how negative and positive financial performance gaps motivate organizations to build more resilient systems, we develop a conceptual process model to reveal the process by which financially and sustainability‐driven organizations can translate these negative and positive financial performance gaps into organizational resilience. We specify the different modes of search behaviors that these organizations pursue when encountering negative and positive financial performance gaps. We then expand on group engagement model to theorize that vicarious search is likely to encourage limiting behaviors, whereas internal search is likely to foster promotion behaviors. Finally, we explain how both promoting and limiting behaviors can be helpful in improving organizational resilience. In this way, we hope to advance research that connects and integrates relatively disparate realms and, more specifically, to contribute to the sustainability, resilience, and performance feedback literatures.     

Pains and gains of environmental management system certification for the sustainable development of manufacturing companies: Heterogeneous effects of industry peer learning

Environmental sustainability is a growing global concern. Environmental management systems (EMS) could be an effective strategic tool to help firms deal with their sustainable development. However, whether EMS certification pays off financially and how it takes effect can be debated. Thus far, these questions remain largely under‐researched. In particular, the effects of EMS certification on financial performance are inconclusive, and the reasons explaining the effects are underdeveloped. This study aims to enrich the current research by exploring the mediating and moderating roles from the perspective of cost‐efficiency trade‐offs to reveal how EMS certification affects financial performance. Applying a PROCESS procedure analysis and causal mediation analysis to a sample of 1,751 Chinese listed manufacturing firms from 2008 to 2016, this study shows that the effect of EMS certification on firms' financial performance is insignificant because their operating costs burden increases while their marketing efficiency and managerial efficiency improve. For the first time, this study demonstrates the moderating role of industry peer learning, as the mediating effects decrease with the growth of industry peer learning.     

Market drivers of sustainability and sustainability learning capabilities: The moderating role of sustainability control systems

This study examines (i) the impact of market drivers of sustainability on the adoption of sustainability learning capabilities and (ii) the moderating role of sustainability control systems (SCS) on the relationship between market drivers of sustainability and sustainability learning capabilities. Drawing on the levers of control framework, stakeholder theory and organisational learning literature, survey data were collected from 175 large scale local and multinational companies operating in Sri Lanka. Findings reveal that market drivers of sustainability have a significant positive impact on sustainability learning capabilities. Whereas the interactive use of SCS shows a positive moderating impact, the diagnostic use of SCS shows a negative impact. The study enhances our understanding of (i) the influence of market drivers of sustainability on the adoption of sustainability learning capabilities and (ii) the use of SCS in enabling sustainability learning capabilities. The study reveals novel insights for managers responding to changing market drivers of sustainability, on how to (re)align different uses of SCS to enable sustainability learning capabilities.     

基于深度学习的跳频信号识别

针对跳频信号分选存在人工提取参数特征具有复杂性的问题，提出了一种基于深度学习的识别方法。首先对跳频信号进行短时傅里叶变换，得到二维的时频矩阵；接着提取信号的轮廓特征，构造三维矩阵作等高线图，并对等高线图进行预处理；最后把预处理后的等高线图输入到卷积神经网络中进行训练、测试，进而实现分类识别。仿真结果表明，在不需要复杂的人工提取参数特征的基础上，在分选率为100〖WT《Times New Roman》〗%〖WTBZ〗时，所提方法经裁剪处理下的信噪比为-15 dB，比支持向量机和传统K-Means聚类算法都低10 dB。实测数据的算法验证表明，所提方法能够将大疆精灵4Pro、hm无人机、司马航模X8HW以及大疆悟2这四类无人机正确分类。     

一种加权聚类划分决策树算法

针对斜划分决策树算法普遍存在时间效率低、部分算法仅能应用于二分类问题，提出了一种基于加权距离的聚类决策树算法。通过Relief-F算法为预测属性计算权重，并将权重用于树结点中数据的聚类过程，使用分簇结果对结点进行多路划分，得到可直接用于多分类问题的决策树。理论分析和实验结果表明，该算法与经典轴平行决策树相比，拥有更好的泛化能力以及相近的算法时间复杂度，与大部分斜决策树相比，在付出更少计算代价的前提下，获得了近似的正确率以及模型简洁度。