Search results

The purpose of this study is to explore the potential of trainable activation functions to enhance the performance of deep neural networks, specifically ResNet architectures, in the task of image classification. By introducing activation functions that adapt during training, the authors aim to determine whether such flexibility can lead to improved learning outcomes and generalization capabilities compared to static activation functions like ReLU. This research seeks to provide insights into how dynamic nonlinearities might influence deep learning models' efficiency and accuracy in handling complex image data sets.

This research integrates three novel trainable activation functions – CosLU, DELU and ReLUN – into various ResNet-n architectures, where “n” denotes the number of convolutional layers. Using CIFAR-10 and CIFAR-100 data sets, the authors conducted a comparative study to assess the impact of these functions on image classification accuracy. The approach included modifying the traditional ResNet models by replacing their static activation functions with the trainable variants, allowing for dynamic adaptation during training. The performance was evaluated based on accuracy metrics and loss profiles across different network depths.

The findings indicate that trainable activation functions, particularly CosLU, can significantly enhance the performance of deep learning models, outperforming the traditional ReLU in deeper network configurations on the CIFAR-10 data set. CosLU showed the highest improvement in accuracy, whereas DELU and ReLUN offered varying levels of performance enhancements. These functions also demonstrated potential in reducing overfitting and improving model generalization across more complex data sets like CIFAR-100, suggesting that the adaptability of activation functions plays a crucial role in the training dynamics of deep neural networks.

This study contributes to the field of deep learning by introducing and evaluating the impact of three novel trainable activation functions within widely used ResNet architectures. Unlike previous works that primarily focused on static activation functions, this research demonstrates that incorporating trainable nonlinearities can lead to significant improvements in model performance and adaptability. The introduction of CosLU, DELU and ReLUN provides a new pathway for enhancing the flexibility and efficiency of neural networks, potentially setting a new standard for future deep learning applications in image classification and beyond.

The purpose of this study is to enrich the literature on employee relations with a new model focusing on the effect of institutional structure and that of employees’ organizational identification on the relationship between institutional structure in an organization and employees’ pro-environmental behaviors, which represents an alternative approach for understanding employees’ pro-environmental performance.

We collect multi-level and multi-source data from 52 four- or five-star hotels in China (N = 963). For data analysis, we adopt the approach of multilevel structural equation modeling.

The results suggest that organizations’ green institutional structure (G-structure) can significantly influence employees’ organizational identification, which in turn can increase their pro-environmental performance.

We propose a new multi-level theoretical perspective to explain employees’ pro-environmental behaviors. While prior studies on the issue mainly consider only the effects of such micro-level variables as ability, motivation and personality, we focus on the effect of organizational institution and its interaction with micro-level variables so that we can evaluate the effect a commonly-studied contextual variable, i.e. green institutions, on the behaviors. Moreover, in this new theoretical model, we also take into account the effect of another insufficiently-tested micro-level variable, i.e. employees’ identification, which has not been considered as frequently as other micro-level variables in studying employees’ pro-environmental performance. Our results highlight the importance of all these variables and suggest a valuable alternative model for more comprehensive research of employees’ green performance.

This study aims to develop a green, economical and efficient ultrasonic-/microwave assisted extraction (UMAE) process for the extraction of anthocyanins.

After optimizing the extraction conditions by response surface methodology, three assays including DPPH, ABTS·+, FRAP were applied to analyze the antioxidant activity of the extracted anthocyanins. The stability under different temperatures, reductant concentrations and pHs was also discussed. The components of anthocyanins in blueberry were analyzed by HPLC-QTOF-MS2.

The optimal extraction parameters were ultrasonic power of 300 W, microwave power of 365.28 W and solid–liquid ratio of 30 (g/mL). The possible structures can be speculated as Delphinidin-3-O-galactoside, Delphinidin, Petunidin, Delphinidin-3-O-glucoside, Petunidin-3-O-glucoside, Cyanidin-3-O-glucoside. The results demonstrated that the UMAE can improve the yield of anthocyanins in shorter extraction time with higher activity.

The present study may provide a promising and feasible route for extracting anthocyanins from blueberries and studying their physicochemical properties, ultimately promoting the utilization of blueberry anthocyanins.

Based on theory Z of leadership, this research aims to investigate the direct relationship between participative leadership (PL) and creative idea validation (CIV) fostering both radical (RC) and incremental creativity (IC). Additionally, by utilizing attribution theory, it explores the moderating effect of coworkers’ knowledge-sharing behavior (KSB) on both the direct and indirect relationships.

Data were collected through a descriptive approach and convenient sampling from three sources – leaders, subordinates and coworkers – in R&D departments at multi-levels within 97 high-tech firms in China. Data comprised 446 employees (subordinates and coworkers) and 94 leaders, and multilevel path analysis was conducted using Mplus software.

The results indicate that PL exhibits both a direct and indirect positive association with RC and IC through the CIV. Moreover, the relationship is enhanced by coworkers’ high-KSB.

Our study offers implications that managers can leverage to foster employee creativity. Leaders are encouraged to embrace a PL style for collective idea validation. However, to overcome coworkers’ reciprocal behavior, they may facilitate trust and team-building exercises, enabling employees to strengthen relationships and share critical information and knowledge resources for the development and validation.

This study is the first to empirically extend the relationship between PL and CIV, utilizing a multilevel approach to assess its impact on distinctive types of creativity – namely, radical and incremental. Further, it testifies the significance of coworkers’ knowledge as an attribution effect influencing the relationships.

Considering the role of analysts in disseminating information, the paper explains the idiosyncratic volatility puzzle of China's stock market. As the largest developing country, China's research can provide meaningful reference for the research of financial markets in other new countries.

From the perspective of behavior, establishing a direct link between individual investor attention and stock price overvaluation.

The authors find that there is a significant idiosyncratic volatility puzzle in China's stock market. Due to the role of mispricing, individual investor attention significantly enhances the idiosyncratic volatility effect, that is, as individual investor attention increases, the greater the idiosyncratic volatility, the lower the expected return. Attention can explain the idiosyncratic volatility puzzle in China's stock market. In addition, due to the role of information production and dissemination, securities analysts can reduce the degree of market information asymmetry and enhance the transparency of market information.

China is the second largest economy in the world, and few scholars analyze it from the perspective of investors' attention. The authors believe this paper has the potential in contributing to the academia.

The purpose of this research study is to improve sentiment analysis (SA) at the aspect level, which is accomplished through two independent goals of aspect term and opinion extraction and subsequent sentiment classification.

The proposed architecture uses neighborhood and dependency tree-based relations for target opinion extraction, a domain–ontology-based knowledge management system for aspect term extraction, and deep learning techniques for classification.

The authors use different deep learning architectures to test the proposed approach of both review and aspect levels. It is reported that Vanilla recurrent neural network has an accuracy of 83.22%, long short-term memory (LSTM) is 89.87% accurate, Bi-LSTM is 91.57% accurate, gated recurrent unit is 65.57% accurate and convolutional neural network is 82.33% accurate. For the aspect level analysis, ρ_aspect comes out to be 0.712 and Δ²_aspect is 0.384, indicating a marked improvement over previously reported results.

This study suggests a novel method for aspect-based SA that makes use of deep learning and domain ontologies. The use of domain ontologies allows for enhanced aspect identification, and the use of deep learning algorithms enhances the accuracy of the SA task.

This study aims to establish a model for rapid and accurate emotion recognition in restaurant online reviews, thus advancing the literature and providing practical insights into electronic word-of-mouth management for the industry.

This study elaborates a hybrid model that integrates deep learning (DL) and a sentiment lexicon (SL) and compares it to five other models, including SL, random forest (RF), naïve Bayes, support vector machine (SVM) and a DL model, for the task of emotion recognition in restaurant online reviews. These models are trained and tested using 652,348 online reviews from 548 restaurants.

The hybrid approach performs well for valence-based emotion and discrete emotion recognition and is highly applicable for mining online reviews in a restaurant setting. The performances of SL and RF are inferior when it comes to recognizing discrete emotions. The DL method and SVM can perform satisfactorily in the valence-based emotion recognition.

These findings provide methodological and theoretical implications; thus, they advance the current state of knowledge on emotion recognition in restaurant online reviews. The results also provide practical insights into intelligent service quality monitoring and electronic word-of-mouth management for the industry.

This study proposes a superior model for emotion recognition in restaurant online reviews. The methodological framework and steps are elucidated in detail for future research and practical application. This study also details the performances of other commonly used models to support the selection of methods in research and practical applications.

This study aims to explore the adverse impacts of abusive supervision on helping behaviors among employees, as mediating by intention to leave and moderating by Islamic work ethics (IWE).

A quantitative approach was employed, and the sample consisted of 283 nurses working in various public sector hospitals in Pakistan. The data analysis was conducted using SPSS and AMOS with the PROCESS macro.

The results suggest that abusive supervision diminishes helping behavior among nurses. Additionally, the study reveals that intention to leave mediates the relationship of abusive supervision and nurses' helping behavior. Moreover, the introduction of IWE as a boundary condition reveals that the mediated link is weaker when IWE is higher, and vice versa.

This study provides valuable insights for hospital authorities to develop intervention strategies and policies aimed at reducing abusive supervision in hospitals. Hospital management should also be aware of the detrimental effects of abusive supervision on nurses' helping behaviors, which can be mitigated by promoting ethical values aligned with IWE.

This study makes a valuable contribution to the limited research on the link between abusive supervision and helping behaviors in hospital settings. It offers new perspectives by incorporating the Conservation of Resources theory, particularly within the healthcare sector. Furthermore, this research expands the current knowledge by investigating the mediating influence of intention to leave and the moderating effect of IWE in mitigating the adverse impact of abusive supervision on nurses' helping behavior in Pakistan's public sector hospitals.