Search results

The purpose of this paper is to alleviate the problem of poor robustness and over-fitting caused by large-scale data in collaborative filtering recommendation algorithms.

Interpreting user behavior from the probabilistic perspective of hidden variables is helpful to improve robustness and over-fitting problems. Constructing a recommendation network by variational inference can effectively solve the complex distribution calculation in the probabilistic recommendation model. Based on the aforementioned analysis, this paper uses variational auto-encoder to construct a generating network, which can restore user-rating data to solve the problem of poor robustness and over-fitting caused by large-scale data. Meanwhile, for the existing KL-vanishing problem in the variational inference deep learning model, this paper optimizes the model by the KL annealing and Free Bits methods.

The effect of the basic model is considerably improved after using the KL annealing or Free Bits method to solve KL vanishing. The proposed models evidently perform worse than competitors on small data sets, such as MovieLens 1 M. By contrast, they have better effects on large data sets such as MovieLens 10 M and MovieLens 20 M.

This paper presents the usage of the variational inference model for collaborative filtering recommendation and introduces the KL annealing and Free Bits methods to improve the basic model effect. Because the variational inference training denotes the probability distribution of the hidden vector, the problem of poor robustness and overfitting is alleviated. When the amount of data is relatively large in the actual application scenario, the probability distribution of the fitted actual data can better represent the user and the item. Therefore, using variational inference for collaborative filtering recommendation is of practical value.

This paper aims to study the effect of temperature on the process and kinetic parameters of the hydrogen evolution reaction of X80 under cathodic protection (CP) in 3.5% NaCl solution.

Potentiodynamic polarization combined with the hydrogen permeation test is used to analyze the hydrogen evolution reaction (HER) process and the rate-determining step for which is diagnosed through the electrochemical impedance spectrum method. Then, the influence of temperature on kinetic parameters of HER can be known from the results obtained by using the Iver-Pickering-Zamenzadeh model for data analysis.

The results show that the HER proceeds through Volmer–Tafel route with the Volmer reaction acting as the rate-controlling step; Increasing temperature gives a higher activity of the HER on X80, it also accelerates the hydrogen desorption and diffusion of hydrogen into the metal.

There exist few studies on the topic of how temperature affects the HER process. It is imperative to conduct a relevant study to give some instruction in cathodic protection system design and this paper fulfills this need.

The purpose of this paper is to unify the expression of fractional grey accumulating generation operator and the reducing generation operator, and build the FDGM(1,1) model with the unified fractional grey generation operator.

By systematically studying the properties of the fractional accumulating operator and the reducing operator, and analyzing the sensitivity of the order value, a unified expression of the fractional operators is given. The FDGM(1,1) model with the unified fractional grey generation operator is established. The relationship between the order value and the modeling error distribution is studied.

The expression of the fractional accumulating generation operator and the reducing generation operator can be unified to a simple expression. For −1<r < 1, the fractional grey generation operator satisfies the principle of new information priority. The DGM(1,1) model is a special case of the FDGM(1,1) model with r = 1.

The sensitivity of the unified operator is verified through random numerical simulation method, and the theoretical proof was not yet possible.

The FDGM(1,1) model has a higher modeling accuracy and modeling adaptability than the DGM(1,1) by optimizing the order.

The expression of the fractional accumulating generation operator and the reducing generation operator is firstly unified. The FDGM(1,1) model with the unified fractional grey generation operator is firstly established. The unification of the fractional accumulating operator and the reducing operator improved the theoretical basis of grey generation operator.

The purpose of this paper is to analyze the mechanism and filter efficacy of accumulation generation operator (AGO)/inverse accumulation generation operator (IAGO) in the frequency domain.

The AGO/IAGO in time domain will be transferred to the frequency domain by the Fourier transform. Based on the consistency of the mathematical expressions of the AGO/IAGO in the gray system and the digital filter in digital signal processing, the equivalent filter model of the AGO/IAGO is established. The unique methods in digital signal processing systems “spectrum analysis” of AGO/IAGO are carried out in the frequency domain.

Through the theoretical study and practical example, benefit of spectrum analysis is explained, and the mechanism and filter efficacy of AGO/IAGO are quantitatively analyzed. The study indicated that the AGO is particularly suitable to act on the system's behavior time series in which the long period parts is the main factor. The acted sequence has good effect of noise immunity.

The AGO/IAGO has a wonderful effect on the processing of some statistical data, e.g. most of the statistical data related to economic growth, crop production, climate and atmospheric changes are mainly affected by long period factors (i.e. low-frequency data), and most of the disturbances are short-period factors (high-frequency data). After processing by the 1-AGO, its high frequency content is suppressed, and its low frequency content is amplified. In terms of information theory, this two-way effect improves the signal-to-noise ratio greatly and reduces the proportion of noise/interference in the new sequence. Based on 1-AGO acting, the information mining and extrapolation prediction will have a good effect.

The authors find that 1-AGO has a wonderful effect on the processing of data sequence. When the 1-AGO acts on a data sequence X, its low-pass filtering effect will benefit the information fluctuations removing and high-frequency noise/interference reduction, so the data shows a clear exponential change trends. However, it is not suitable for excessive use because its equivalent filter has poles at the non-periodic content. But, because of pol effect at zero frequency, the 1-AGO will greatly amplify the low-frequency information parts and suppress the high-frequency parts in the information at the same time.

From the intra- and inter-regional dimensions, this paper investigates the linkage between industrial agglomeration and inventory performance, and further demonstrates the moderating role of firm size and enterprise status in the supply chain on this linkage.

Using a large panel dataset of Chinese manufacturers in the Yangtze River Delta for the period from 2008 to 2013, this study employs the method of spatial econometric analysis via a spatial Durbin model (SDM) to examine the effects of industrial agglomeration on inventory performance. Meanwhile, the moderation model is applied to examine the moderating role of two firm-level heterogeneity factors.

At its core, this research demonstrates that industrial agglomeration is associated with the positive change of inventory performance in the adjacent regions, whereas that in the host region as well as in general does not significantly increase. Additionally, both firm size and enterprise status in the supply chain can positively moderate these effects, except for the moderating role of firm size on the positive spillovers.

In view of firm heterogeneity, managers should take special care when matching their abilities of inventory management with the agglomeration effects. Firms with a high level of inventory management are suited to stay in an industrial cluster, while others would be better in the adjacent regions to enhance inventory performance.

This paper is the first to systematically analyze the effects of industrial agglomeration on inventory performance within and across clusters, and confirm that these effects are contingent upon firm size and enterprise status in the supply chain. It adds to the existing literature by highlighting the spatial spillovers from industrial clusters and enriching the antecedents of inventory leanness.

This paper explores the relationship between inventory productivity and the likelihood of venture survival and then examines how financial constraints moderate the inventory productivity–survival linkage.

Accelerated failure time (AFT) model is employed to study the link between inventory productivity and venture survival by using small- and medium-sized enterprise (SME) data from Chinese Annual Survey of Industrial Firms (CASIF) database over the period 1999–2007.

The paper demonstrates a converse U-curve relation between inventory productivity and venture survival. Additionally, financial constraints as the moderator weaken the marginal effect of inventory productivity on venture survival.

Managers should pay more attention to the important inventory performance indicator: inventory productivity. In the context of prominent financing difficulties, managers should be rapid to adjust the competitive strategy and optimize the internal production process according to the inherent nature of risks in a friction environment, and thus generate resources that enterprises cannot raise in the financial market.

This study may be the first to practically investigate the role of inventory productivity on venture survival and the moderating effect of financing constraints on this relationship. It adds to abundant articles as regards the interface between operation management and venture survival by exploring how financial constraints moderate the inventory productivity–survival linkage.

Smog seriously affects the ecological environment and poses a threat to public health. Therefore, smog control has become a key task in China, which requires reliable prediction.

This paper establishes a novel time-lag GM(1,N) model based on interval grey number sequences. Firstly, calculating kernel and degree of greyness of the interval grey number sequence respectively. Then, establishing the time-lag GM(1,N) model of kernel and degree of greyness sequences respectively to obtain their values after determining the time-lag parameters of two models. Finally, the upper and lower bounds of interval grey number sequences are obtained by restoring the values of kernel and degree of greyness.

In order to verify the validity and practicability of the model, the monthly concentrations of PM2.5, SO2 and NO2 in Beijing during August 2017 to September 2018 are selected to establish the time-lag GM(1,3) model for kernel and degree of greyness sequences respectively. Compared with three existing models, the proposed model in this paper has better simulation accuracy. Therefore, the novel model is applied to forecast monthly PM2.5 concentration for October to December 2018 in Beijing and provides a reference basis for the government to formulate smog control policies.

The proposed model can simulate and forecast system characteristic data with the time-lag effect more accurately, which shows that the time-lag GM(1,N) model proposed in this paper is practical and effective.

Based on interval grey number sequences, the traditional GM(1,N) model neglects the time-lag effect of driving terms, hence this paper introduces the time-lag parameters into driving terms of the traditional GM(1,N) model and proposes a novel time-lag GM(1,N) model.

This study aims to develop a data fusion method for powder-bed fusion (PBF) process monitoring based on process image information. The data fusion method can help improve process condition identification performance, which can provide guidance for further PBF process monitoring and control system development.

Design of reliable process monitoring systems is an essential approach to solve PBF built quality. A data fusion framework based on support vector machine (SVM), convolutional neural network (CNN) and Dempster-Shafer (D-S) evidence theory are proposed in the study. The process images which include the information of melt pool, plume and spatters were acquired by a high-speed camera. The features were extracted based on an appropriate image processing method. The three feature vectors corresponding to the three objects, respectively, were used as the inputs of SVM classifiers for process condition identification. Moreover, raw images were also used as the input of a CNN classifier for process condition identification. Then, the information fusion of the three SVM classifiers and the CNN classifier by an improved D-S evidence theory was studied.

The results demonstrate that the sensitivity of information sources is different for different condition identification. The feature fusion based on D-S evidence theory can improve the classification performance, with feature fusion and classifier fusion, the accuracy of condition identification is improved more than 20%.

An improved D-S evidence theory is proposed for PBF process data fusion monitoring, which is promising for the development of reliable PBF process monitoring systems.

The purpose of this paper is to investigate the impact of the nature of firm heterogeneity and factors of the host country on the choice of entry modes in greenfield investments and cross-border mergers and acquisitions.

An empirical analysis was conducted of 450 outward foreign direct investment (OFDI) cases of Chinese-listed companies from 2001 to 2015. A regression analysis was conducted to determine the influence of the heterogeneous nature of enterprises and host country factors on the choice of entry mode.

First, the nature of a firm’s heterogeneity differs in terms of their mobile or immobile capabilities, which may affect entry strategies. Second, although Chinese multinational companies do not have a strong ownership advantage when compared with multinational companies in developed countries, they have certain marketing capabilities, such as innovations, aimed at customer needs that make it possible to implement their internationalization strategy. Third, factors such as cultural distance and investment risk of the host country significantly influence the choice of OFDI entry modes.

The authors discuss the mobility of a firm’s resource heterogeneity in determining Chinese firms’ entry mode choices and emphasize that Chinese marketing-intensive firms seek complementary resources from the firms of the host countries to achieve competitive advantages. The authors further divide heterogeneous enterprise resources into research and development resources and marketing resources according to the degree of international mobility and examine what kind of firm heterogeneity could help in the selection of different entry modes.

The purpose of this paper is to analyze the spectral characteristics of moving average operator and to propose a novel time-frequency hybrid sequence operator.

Firstly, the complex data is converted into frequency domain data by Fourier transform. An appropriate frequency domain operator is constructed to eliminate the impact of disturbance. Then, the inverse Fourier transform transforms the frequency domain data in which the disturbance is removed, into time domain data. Finally, an appropriate moving average operator of N items is selected based on spectral characteristics to eliminate the influence of periodic factors and noise.

Through the spectrum analysis of the real-time data sensed and recorded by microwave sensors, the spectral characteristics and the ranges of information, noise and shock disturbance factors in the data can be clarified.

The real-time data analysis results for a drug component monitoring show that the hybrid sequence operator has a good effect on suppressing disturbances, periodic factors and noise implied in the data.

Firstly, the spectral analysis of moving average operator and the novel time-frequency hybrid sequence operator were presented in this paper. For complex data, the ideal effect is difficult to achieve by applying the frequency domain operator or time domain operator alone. The more satisfactory results can be obtained by time-frequency hybrid sequence operator.