Search results

A lot of triaxial compressive models for different concrete types and different concrete strength classes were proposed to be used in structural analyses. The existence of so many models creates conflicts and confusions during the selection of the models. In this study, reliability analyses were carried out to prevent such conflicts and confusions and to determine the most reliable model for normal- and high-strength concrete (NSC and HSC) under combined triaxial compressions. The paper aims to discuss these issues.

An analytical model was proposed to estimate the strength of NSC and HSC under different triaxial loadings. After verifying the validity of the model by making comparisons with the models in the literature, reliabilities of all models were investigated. The Monte Carlo simulation method was used in the reliability studies. Artificial experimental data required for the Monte Carlo simulation method were generated by using artificial neural networks.

The validity of the proposed model was verified. Reliability indexes of triaxial compressive models were obtained for the limit states, different concrete strengths and different lateral compressions. Finally, the reliability indexes were tabulated to be able to choose the best model for NSC and HSC under different triaxial compressions.

Concrete compressive strength and lateral compression were taken as variables in the model.

The reliability indexes were tabulated to be able to choose the best model for NSC and HSC under different triaxial compressions.

A new analytical model was proposed to estimate the strength of NSC and HSC under different triaxial loadings. Reliability indexes of triaxial compressive models were obtained for the limit states, different concrete strengths and different lateral compressions. Artificial experimental data were obtained by using artificial neural networks. Four different artificial neural networks were developed to generate artificial experimental data. They can also be used in the estimations of the strength of NSC and HSC under different triaxial loadings.

This study explores challenges facing the applicability of deep learning (DL) in software-defined networks (SDN) based campus networks. The study intensively explains the automation problem that exists in traditional campus networks and how SDN and DL can provide mitigating solutions. It further highlights some challenges which need to be addressed in order to successfully implement SDN and DL in campus networks to make them better than traditional networks.

The study uses a systematic literature review. Studies on DL relevant to campus networks have been presented for different use cases. Their limitations are given out for further research.

Following the analysis of the selected studies, it showed that the availability of specific training datasets for campus networks, SDN and DL interfacing and integration in production networks are key issues that must be addressed to successfully deploy DL in SDN-enabled campus networks.

This study reports on challenges associated with implementation of SDN and DL models in campus networks. It contributes towards further thinking and architecting of proposed SDN-based DL solutions for campus networks. It highlights that single problem-based solutions are harder to implement and unlikely to be adopted in production networks.

The purpose of this study is to make a prediction of the R&D output of China from the perspective of R&D institutions and put forward a set of policy recommendations for further development of the science and technology in China.

In this paper, an improved discrete grey multivariable model is proposed, which takes both the interaction effects and the accumulative effects into account. As the current research on China's R&D activities is generally based on the perspective of universities or industrial enterprises above designated size while few studies put their focus on R&D institutions, this paper applies the proposed model to carry out an empirical analysis based on the data of China's R&D institutions from 2009 to 2019. The prediction results from the new model are then compared with three existing approaches and the comparison results indicate that the proposed model generally outperforms existing methods. A further prediction of the R&D output in China's R&D institutions is conducted into a future horizon from 2020 to 2023 by using the model.

The results indicate that China's R&D institutions have a good development trend and broad prospects, which is closely related to China's long-term investment in science and technology. Additionally, the R&D inputs of China possess obvious interaction effects and accumulative effects.

The paper considers the interaction effects and the accumulative effects of R&D inputs simultaneously and proposes an improved discrete grey multivariable model, which fills the gap in previous studies.

The aim of this study is to examine the tracking performance and tracking error (TE) of New Zealand exchange traded fsunds (ETFs).

The authors use regression methods and cointegration analysis to examine tracking performance. Multivariate regressions are used to examine the determinants of TE.

At the daily frequency, the authors observe that the ETFs have substantially different exposures to their underlying indexes from what they should be, which is confirmed by cointegration analysis. At the monthly frequency, tracking performance improves but still shows significant differences between the ETF and its underlying index. When the authors examine the TEs of the ETFs, the authors observe that these are substantial and that there is considerable variation in TE. Regression analysis shows that both characteristics of the ETF and the constituents of the index the ETF tracks, as well as the volatility of the underlying benchmark are determinants of the TE of the ETFs.

This is the first study to examine New Zealand-based ETFs. The findings contribute to understanding the performance of these ETFs and are of relevance to academics, investors and the ETF provider.

User feedback inferred from the user's search-time behavior could improve the learning to rank (L2R) algorithms. Click models (CMs) present probabilistic frameworks for describing and predicting the user's clicks during search sessions. Most of these CMs are based on common assumptions such as Attractiveness, Examination and User Satisfaction. CMs usually consider the Attractiveness and Examination as pre- and post-estimators of the actual relevance. They also assume that User Satisfaction is a function of the actual relevance. This paper extends the authors' previous work by building a reinforcement learning (RL) model to predict the relevance. The Attractiveness, Examination and User Satisfaction are estimated using a limited number of the features of the utilized benchmark data set and then they are incorporated in the construction of an RL agent. The proposed RL model learns to predict the relevance label of documents with respect to a given query more effectively than the baseline RL models for those data sets.

In this paper, User Satisfaction is used as an indication of the relevance level of a query to a document. User Satisfaction itself is estimated through Attractiveness and Examination, and in turn, Attractiveness and Examination are calculated by the random forest algorithm. In this process, only a small subset of top information retrieval (IR) features are used, which are selected based on their mean average precision and normalized discounted cumulative gain values. Based on the authors' observations, the multiplication of the Attractiveness and Examination values of a given query–document pair closely approximates the User Satisfaction and hence the relevance level. Besides, an RL model is designed in such a way that the current state of the RL agent is determined by discretization of the estimated Attractiveness and Examination values. In this way, each query–document pair would be mapped into a specific state based on its Attractiveness and Examination values. Then, based on the reward function, the RL agent would try to choose an action (relevance label) which maximizes the received reward in its current state. Using temporal difference (TD) learning algorithms, such as Q-learning and SARSA, the learning agent gradually learns to identify an appropriate relevance label in each state. The reward that is used in the RL agent is proportional to the difference between the User Satisfaction and the selected action.

Experimental results on MSLR-WEB10K and WCL2R benchmark data sets demonstrate that the proposed algorithm, named as SeaRank, outperforms baseline algorithms. Improvement is more noticeable in top-ranked results, which usually receive more attention from users.

This research provides a mapping from IR features to the CM features and thereafter utilizes these newly generated features to build an RL model. This RL model is proposed with the definition of the states, actions and reward function. By applying TD learning algorithms, such as the Q-learning and SARSA, within several learning episodes, the RL agent would be able to learn how to choose the most appropriate relevance label for a given pair of query–document.

Marketing is a hot topic, and the purpose of this study is to investigate how shareholding strategies can be applied to achieve strategic synergy between firms in vertical supply chains to improve retailers’ marketing efforts from a long-term perspective.

This study constructs Stackelberg models to analyze the operating mechanisms of shareholding supply chains under forward, backward and cross-shareholding strategies. The authors analyze the effects of shareholding on prices, marketing efforts and profits, and explore the strategic preferences and outcomes of different supply chain members.

Forward/backward shareholding plays the same role as cross/nonshareholding in supply chains because the effect of the retailer’s shareholding is offset by the power status of the manufacturer, and the retailer can still profit when wholesale prices are higher than selling prices in certain cases. A manufacturer’s shareholding in a retailer can benefit consumers and improve marketing efforts by reducing retailers’ marketing costs, while a retailer’s shareholding in a manufacturer has no such effect. None of all shareholding strategies can coordinate the interests of all members; however, an effective rebate policy can resolve this problem.

The results reveal the operational mechanism of shareholding supply chains and provide reference values for managers who want to improve marketing efforts and economic performance using a shareholding strategy.

This paper considers carbon abatement in a competitive supply chain that is composed of a manufacturer and two retailers under vertical shareholding. The authors emphasize the equilibrium decision problem of stakeholders under vertical shareholding and different power structures.

A game-theoretic approach was used to probe the influence of power structure and retailer competition on manufacturers' carbon abatement under vertical shareholding. The carbon abatement decisions, environmental imp4cacts (EIs) and social welfare (SW) of different scenarios under vertical shareholding are obtained.

The findings show that manufacturers are preferable to carbon abatement and capture optimal profits when shareholding is above a threshold under the retailer power equilibrium, but they may exert a worse negative impact on the environment. The dominant position of the held retailer is not always favorable to capturing the optimal SW and mitigating EIs. In addition, under the combined effect of competition level and shareholding, retailer power equilibrium scenarios are more favorable to improving SW and reducing EIs.

This paper inspects the combined influence of retailer competition and power structure on manufacturers' carbon abatement. Distinguishing from previous literature, the authors also consider the impact of vertical shareholding and consumer preferences. In addition, the authors analyze the SW and EIs in different scenarios.

The purpose of this paper is to solve the problem that the qualitative relative factors cannot be employed in traditional multivariate grey models.

First, a new model is constructed though introducing dummy drivers. Then, the parameters estimation method and recursive function of the model are discussed. Furthermore, dummy driver setting, pre and post test methods of dummy drivers are proposed. At last, the per capita income forecasting of rural residents in Henan province of China is solved with the proposed model.

The proposed model is the reasonable extension of original one. The accuracy of it is higher than former model. In the case study, the forecasting results of proposed model are compared with other grey forecasting models, and prove that proposed model has not only high accuracy, but also clear physical meaning.

The method proposed in the paper could be used in policy effect measure, marketing forecasting, etc., when the predictor variables are influenced by some qualitative variables.

It will promote the accuracy of multivariate grey forecasting model.

In the face of complex and challenging economic and business environments, developing and implementing approaches to predict bankruptcy has become important for firms. Bankruptcy prediction can be regarded as a grey system problem because while factors such as the liquidity, solvency and profitability of a firm influence whether it goes bankrupt, the precise manner in which these factors influence the discrimination between failed and non-failed firms is uncertain. In view of the applicability of multivariate grey prediction models (MGPMs), this paper aimed to develop a grey bankruptcy prediction model (GBPM) based on the GM (1, N) (BP-GM (1, N)).

As the traditional GM (1, N) is designed for time series forecasting, it is better to find an appropriate permutation of firms in the financial data as if the resulting sequences are time series. To solve this challenging problem, this paper proposes GBPMs by integrating genetic algorithms (GAs) into the GM (1, N).

Experimental results obtained for the financial data of Taiwanese firms in the information technology industries demonstrated that the proposed BP-GM (1, N) performs well.

Among artificial intelligence (AI)-based techniques, GBPMs are capable of explaining which of the financial ratios has a stronger impact on bankruptcy prediction by driving coefficients.

Applying MGPMs to a problem without relation to time series is challenging. This paper focused on bankruptcy prediction, a crucial issue in financial decision-making for businesses, and proposed several GBPMs.