改进的多层感知机在客户流失预测中的应用

针对在传统的客户流失预测数据预处理中，使用one-hot编码处理离散属性导致数据维度增加及数据过于稀疏的问题，提出了两种基于多层感知机的改进后的客户流失预测模型。其主要思想是分别使用堆叠自编码器和实体嵌入两种方法对多层感知机进行改进，通过将离散属性的高维编码数据向低维空间映射，有效地减少了one-hot编码产生的稀疏数据，增加了离散属性值之间的关联度。在对两份公开的数据集进行交叉验证后的实验结果表明，改进后的模型既有效地提高了预测的准确度，又维持了传统多层感知机模型在并行化计算方面的优势。     

How vloggers embrace their viewers: Focusing on the roles of para-social interactions and flow experience

Social media allow marketers to directly influence consumers through a diverse set of tools. The characteristics of individual platforms enable the development of creative and innovative strategies for social media marketing. This study aims at examining para-social interactions (PSI), their antecedents (i.e., attitude homophily and physical attractiveness), and their consequences (i.e., consumers’ sense of belonging, flow experience, urge to purchase, addiction, and impulse purchase). The mediating role of flow experience is also examined in this study. Findings, implications, and directions for further research are also presented in this study.     

The pollution concern in the era of globalization: Do the contribution of foreign direct investment and trade openness matter?

This analysis aims to treat the contribution of the foreign direct investment inflows and trade openness to environmental degradation. Drawing on the data for 27 African countries over the period from 1990 to 2013, we develop an empirical model based on a set of panel methods. This approach includes Westerlund (2007) panel cointegration technic, Common Correlated Effects Mean Group (CCE-MG) estimation procedure, and the Granger causality test. In terms of long-term causality, our findings reveal bidirectional long-term causality between CO₂/NO_x emissions, GDP, trade openness, and foreign direct investment. Besides, in the short-run, the results recorded a unidirectional causality running from GDP to CO₂, and from FDI to CO₂. However, bidirectional causality nexus was detected between trade openness and CO₂. Additionally, the findings portray bidirectional causality between GDP and NO_x and between trade openness and NO_x, while a unidirectional causality from foreign direct investment to NO_x. Our analysis substantiates the importance of the foreign direct investment inflows and trade openness in mitigating the adverse effect of heavy pollutant activities and resolving the environmental puzzle (growing without polluting) consistent with the Millennium Development Goals (MGDs) of the UN background.     

Ex post demand information sharing between differentiated suppliers and a common retailer

We consider ex post demand information sharing in a two-echelon supply chain in which two suppliers sell differentiated products through a common retailer. We model three scenarios of information sharing to characterise the conditions under which information sharing may benefit or hurt different players: (1) no supplier is informed; (2) only one supplier is informed; and (3) both suppliers are informed. We find that the retailer may voluntarily share the low demand information when the product differentiation and demand magnitude satisfy certain conditions. In contrast, the retailer has no incentive to share the high demand information, even though this information benefits the suppliers. To achieve information sharing, we propose a side payment mechanism, which allows benefitted players to subsidise hurtful players. Finally, we extend our model by considering a situation where the suppliers have limited capacity and may invest to ramp up production capacity to satisfy a high demand.     

Investigation of seasonal variations of Beijing pavement condition data using unevenly spaced dynamic panel data model

Beijing municipal highway administration started to collect pavement condition data on its major expressways since 2006. It is advised in the Chinese practice standard that data collection shall be conducted on annual basis. However, pavement data are usually collected at different seasons of the year, which may cause significant seasonal variations in the observed condition. Moreover, for some reasons, data are missing for some of the pavement sections at certain years, which could bring up difficulties in performance model estimation and inference. These concerns have been simply neglected in past practice. This study proposed an unevenly spaced dynamic panel data model to investigate the seasonal patterns of a performance indicator called Ride Quality Index (RQI). A quasi-differencing approach was adopted for the estimation. Data collected from the 5th Ring Road of Beijing were used in the case study. It was found that RQI data collected during the fall season are expected to be lower than that collected during the spring or summer seasons. Findings from this research would be helpful to pavement engineers in using unevenly spaced pavement condition data for future condition estimation.     

Urban landscape as a spatial representation of land rent: A quantitative analysis

Due to the emergence of geographical ‘big data,’ the field of urban studies is enjoying many new research opportunities. By using several sources of geographical ‘big data’, an analysis framework was structured to measure the urban landscape based on three aspects: city plan, pattern of building form, and urban land use. An association rule analysis was used to explore the relationship between land rent and the urban landscape, and the results indicate that the urban landscape differs across urban areas. The blocks classified as being located in main centers were associated with more convenient public transportation, denser road networks, more vertical street space, more diverse block patterns, more flexible architecture arrangements, mixed uses, high density, and more services. By contrast, non-center areas usually comprised blocks that were larger, tabular, single-purpose, and more regular. Non-center areas often cannot provide high-quality public goods, and they contain scattered large industrial enterprises. The urban landscapes of sub-center blocks fell in between these two urban areas. To our knowledge, this is the first paper that attempts to explain the relationship between land rent and urban landscapes based on ‘big data,’ and our study may provide meaningful insights into urban design for government officials and academics.     

Industry effects of oil price shocks: A re-examination

Sectoral responses to oil price shocks help determine how these shocks are transmitted throughout the economy. Textbook treatments of oil price shocks often emphasize negative supply, or cost, effects on oil importing countries. By contrast, the seminal contribution of Lee and Ni (2002) has shown that almost all U.S. industries experience oil price shocks largely through a reduction in their respective demands. Only industries with very high oil intensities face a supply-driven reduction. In this paper, we re-examine this seminal finding using two additional decades of data. Further, we apply updated empirical methods, including structural factor-augmented vector autoregressions that take into account how industries are linked among themselves and with the remainder of the macro-economy. Our results confirm the original finding of Lee and Ni that demand effects of oil price shocks dominate in all but a handful of U.S. industries.     

Influence of market orientation on performance: the moderating roles of customer participation breadth and depth in new product development

ABSTRACT

With a greater number of B2B firms integrating customers into the new product development (NPD) process, how to utilize customer involvement in NPD is an important decision because it may be a double-edged sword carrying both bright and dark sides. Utilizing a sample of 193 B2B firms across various industries, we validate previous research that suggests market orientation positively influences NPD performance and subsequently examine how this relationship may either be enhanced or diminished contingent upon how customers are utilized in the NPD process. The results show that the market orientation–NPD performance relationship is enhanced by having customers participate in a greater number of activities throughout NPD (customer participation breadth) and diminished when customers are involved at deeper levels (customer participation depth). This research suggests that the exact involvement of customers is a critical decision and has clear implications for the dialogues about customer involvement and management of customer relationships.     

Relationship between the oil price volatility and sectoral stock markets in oil-exporting economies: Evidence from wavelet nonlinear denoised based quantile and Granger-causality analysis

This paper examines the extent of volatility between oil price and sectoral indices in the Gulf Cooperation Council (GCC) countries by using quantile regression analysis (QRA) for the return's series and denoised series over the period 2006–2017. Four sectors are found to offer diversification opportunities during a high market (i.e., 90th quantile). All the sectors are found interdependent of oil price volatility; however, the bank and insurance sectors are insusceptible to oil price volatility during the 10th, 25th and 75th quantiles. In addition, QRA results for wavelet nonlinear denoising with a soft-thresholding series indicate that all the sectors are interdependent of oil price volatility but that the aggregate market index, transport and telecommunication sectors are insensitive to oil price volatility during the 75th and 90th quantiles. This highlights the usefulness of denoising the financial returns series when applying regression tools. Moreover, the contagion and interdependence between the oil price and stock returns sectors are estimated by frequency domain causality.     

Smart IoT data platform in hospital and postoperative analgesic effects of orthopedic patients

IOT test this large donor control engineering data can relieve postoperative pain is discussed. One human operation, closed-loop, postoperative analgesic things large data system is designed to alleviate chronic pain. In this system the need of pain relief, when the patient presses the button. Computer interface buttons, the patient's IOT large data after pump. Patients with pain models have been developed to describe the kinetics of chronic pain in people's physiological and psychological reactions. Further including analgesic pharmacokinetic model. Pain model parameters have been adjusted, so that large data things validated self-management of pain behavior mimics a real button, presses the record. Unfortunately, if the adhesion is removed, the inflammation process begins; it is the foundation of a new adhesion causes, creating a vicious cycle difficult to break. Is achieved by electrical stimulation therapy of the small intestine contraction is applied. The results have not been entirely successful, it does not complete the verification of transcutaneous electrical nerve stimulation. However, this idea, a new method is reliable become the treatment of adhesive intestinal obstruction. Effective use of long things big data, as of this site, are, first aid and coherent knowledge of post-operative care, and people to help taking the right medicine for the patient, such research index regret believe that to reduce the occurrence of cans distinguish the analgesic efficacy of the use of different large amounts and compositions of the networking data six hundred and seventy and 9 vertical abdomen, spine and limbs procedures.