The purpose of this paper is to propose a approach for data visualization and industrial process monitoring.
Abstract
Purpose
The purpose of this paper is to propose a approach for data visualization and industrial process monitoring.
Design/methodology/approach
A deep enhanced t-distributed stochastic neighbor embedding (DESNE) neural network is proposed for data visualization and process monitoring. The DESNE is composed of two deep neural networks: stacked variant auto-encoder (SVAE) and a deep label-guided t-stochastic neighbor embedding (DLSNE) neural network. In the DESNE network, SVAE extracts informative features of the raw data set, and then DLSNE projects the extracted features to a two dimensional graph.
Findings
The proposed DESNE is verified on the Tennessee Eastman process and a real data set of blade icing of wind turbines. The results indicate that DESNE outperforms some visualization methods in process monitoring.
Originality/value
This paper has significant originality. A stacked variant auto-encoder is proposed for feature extraction. The stacked variant auto-encoder can improve the separation among classes. A deep label-guided t-SNE is proposed for visualization. A novel visualization-based process monitoring method is proposed.
Details
Keywords
Peipei Lu, Meiping Wu, Xin Liu, Xiaojin Miao and Weipeng Duan
Ti6Al4V is a widely used metal for biomedical application due to its excellent corrosion resistance, biocompatibility and mechanical strength. However, a coupling reaction of…
Abstract
Purpose
Ti6Al4V is a widely used metal for biomedical application due to its excellent corrosion resistance, biocompatibility and mechanical strength. However, a coupling reaction of friction and corrosion is the critical reason for the failure of implants during the long-term service in human body, shortening the life expectancy and clinical efficacy of prosthesis. Hence, this study aims to find a feasible approach to modify the service performances of Ti6Al4V.
Design/methodology/approach
Selective laser melting (SLM), as one of the emerging metal-based additive manufacturing (AM) technologies is capable for fabricating patient-specific personalized customization of artificial prosthesis joints, owing to its high adaptability for complex structures. This study is concerned with the tribocorrosion behavior of SLM fabricated Ti6Al4V substrate enhanced by laser rescanning and graphene oxide (GO) mixing. The tribocorrosion tests were performed on a ball-on-plate configuration under the medium of simulated body fluid (SBF). Moreover, the surface morphologies, microstructures, microhardness and contact angle tests were used to further reveal the in-situ strengthening mechanism of GO/Ti6Al4V nanocomposites.
Findings
The results suggest that the strengthening method of GO mixing and laser rescanning shows its capability to enhance the wear resistance of Ti6Al4V by improving surface morphologies and promoting the generation of hard phases. The wear volume of R-GO/Ti6Al4V is 5.1 × 10−2 mm3, which is 25.0% lower than that of pure SLM-produced Ti6Al4V. Moreover, a wear-accelerated corrosion of the Ti6Al4V occurs in SBF medium, leading to a drop in the open circuit potential (OCP), but R-GO/Ti6Al4V has the lowest tendency to corrosion. Compared to that of pure Ti6Al4V, the microhardness and contact angle of R-GO/Ti6Al4V were increased by 32.89% and 32.60%, respectively.
Originality/value
Previous investigations related to SLM of Ti6Al4V have focused on improving its density, friction and mechanical performances by process optimization or mixing reinforcement phase. The authors innovatively found that the combination of laser rescanning and GO mixing can synergistically enhance the tribocorrosion properties of titanium alloy, which is a feasible way to prolong the service lives of medical implants.
Details
Keywords
Li Dong, Jinlong Chen and Weipeng Wu
This study examines how maturity mismatch, a specific type of financial structure of firms, affects corporate outward foreign direct investment (OFDI).
Abstract
Purpose
This study examines how maturity mismatch, a specific type of financial structure of firms, affects corporate outward foreign direct investment (OFDI).
Design/methodology/approach
Using the number of newly established foreign subsidiaries in a given year as firm-level OFDI and utilizing data from Chinese listed firms between 2007 and 2022, we employ a negative binomial regression model to examine the impact of corporate maturity mismatch on the OFDI. We also make efforts to ensure the robustness of the result, such as employing an exogenous policy to establish a difference-in-difference model.
Findings
The empirical result indicates that maturity mismatch inhibits firms' OFDI. Additional test shows that maturity mismatch increases firms' financing costs and reduces firms' research and development (R&D) investment and that the negative impact of maturity mismatch on OFDI is predominantly observed in firms with high financial constraints and low R&D intensity, indicating that maturity mismatch may affect firms' OFDI through the financing cost channel and the R&D investment channel.
Originality/value
Corporate maturity mismatch is common in China and similar emerging markets. However, research on the economic consequences of maturity mismatch, especially its impact on firms' overseas expansions, is rare. This study establishes the relationship between corporate maturity mismatch and OFDI, contributes to the literature on the relationship between financial factors and OFDI, and provides policy implications for emerging market countries.
Details
Keywords
Xiaofeng Su, Weipeng Zeng, Manhua Zheng, Xiaoli Jiang, Wenhe Lin and Anxin Xu
Following the rapid expansion of data volume, velocity and variety, techniques and technologies, big data analytics have achieved substantial development and a surge of companies…
Abstract
Purpose
Following the rapid expansion of data volume, velocity and variety, techniques and technologies, big data analytics have achieved substantial development and a surge of companies make investments in big data. Academics and practitioners have been considering the mechanism through which big data analytics capabilities can transform into their improved organizational performance. This paper aims to examine how big data analytics capabilities influence organizational performance through the mediating role of dual innovations.
Design/methodology/approach
Drawing on the resource-based view and recent literature on big data analytics, this paper aims to examine the direct effects of big data analytics capabilities (BDAC) on organizational performance, as well as the mediating role of dual innovations on the relationship between (BDAC) and organizational performance. The study extends existing research by making a distinction of BDACs' effect on their outcomes and proposing that BDACs help organizations to generate insights that can help strengthen their dual innovations, which in turn have a positive impact on organizational performance. To test our proposed research model, this study conducts empirical analysis based on questionnaire-base survey data collected from 309 respondents working in Chinese manufacturing firms.
Findings
The results support the proposed hypotheses regarding the direct and indirect effect that BDACs have on organizational performance. Specifically, this paper finds that dual innovations positively mediate BDACs' effect on organizational performance.
Originality/value
The conclusions on the relationship between big data analytics capabilities and organizational performance in previous research are controversial due to lack of theoretical foundation and empirical testing. This study resolves the issue by provides empirical analysis, which makes the research conclusions more scientific and credible. In addition, previous literature mainly focused on BDACs' direct impact on organizational performance without making a distinction of BDAC's three dimensions. This study contributes to the literature by thoroughly introducing the notions of BDAC's three core constituents and fully analyzing their relationships with organizational performance. What's more, empirical research on the mechanism of big data analytics' influence on organizational performance is still at a rudimentary stage. The authors address this critical gap by exploring the mediation of dual innovations in the relationship through survey-based research. The research conclusions of this paper provide new perspective for understanding the impact of big data analytics capabilities on organizational performance, and enrich the theoretical research connotation of big data analysis capabilities and dual innovation behavior.
Details
Keywords
We use disaggregated survey data set to investigate the impact of personality traits on the level of education in the USA. We attempt to shed light on the contribution of each of…
Abstract
Purpose
We use disaggregated survey data set to investigate the impact of personality traits on the level of education in the USA. We attempt to shed light on the contribution of each of the Big Five personality traits on the education decision made by the individuals.
Design/methodology/approach
We use the quantile regression analysis in order to investigate to what extent certain aspects of personality may help an individual to invest in education.
Findings
Our findings uncover a significant effect of noncognitive skills on the level of education. It is shown that people with high emotional stability and agreeableness invest in human capital, especially when we move to the higher quantiles of the conditional distribution function. Moreover, we argue that the estimated signs of the traits remain stable across the quantiles, while the relevant curvatures indicate for the first time in the empirical literature, the presence of nonlinear effects. Last, our model survived robustness checks under the inclusion of two aggregated higher-order factors, namely “Alpha” and “Beta.”
Research limitations/implications
Although we used several control variables (e.g. Gender, Age) to address the impact of noncognitive skills on education, special attention should be given to the use of additional socioeconomic indicators such as the skin color of participants, the urbanization rate, the level of unemployment, the level of income, parental education among others. These measures affect the causality driven by the inclusion of certain economic and demographic characteristics and minimize the endogeneity bias drawn from the inclusion of the sample variables. One additional limitation is that the survey-based data refer only to people with higher education (>13 years of study). Therefore, our empirical findings must be tested on a richer sample to capture the effect of personality traits on a broad spectrum of educational stages (e.g. early learning years, primary education, secondary education, etc.).
Originality/value
Our empirical findings add enough new insights to the existing literature. First, we attempt to assess the role of noncognitive skills proxied by the Big Five Inventory (hereafter “BFI”) on the education decision made by the individuals. Second, we provide fresh evidence of nonlinear effects between personality traits and education totally ignored by the existing literature. Our third contribution is to analyze the role of personality in enhancing the importance of investment in higher education as a determinant of individual behavior. In this way, we contribute to the growing field of behavioral economics since the study of noncognitive skills offers a range of new ideas and expanding research opportunities for social scientists (economists, psychologists, sociologists, etc.).