Search results

Due to the rapid development of artificial intelligence (AI) technology, increasing the use of AI in healthcare is critical, but few studies have explored the extent to which physicians cooperate with AI in their work to achieve productive and innovative performance, which is a key issue in operations management (OM). We conducted empirical research to answer this question.

We developed a conceptual model based on the ambidextrous perspective. To test our model, we collected data from 200 Chinese hospitals. One senior and one junior physician from each hospital participated in this research so that we could get a more comprehensive view. Based on the sample of 400 participants and the conceptual model, we examined whether different types of AI use have distinct impacts on physicians’ productivity and innovation by conducting hierarchical regression and post hoc tests. We also introduced team psychological safety climate (TPSC) and AI technology uncertainty (AITU) as moderators to investigate this topic in further detail.

We found that augmentation AI use is positively related to overall productivity and innovative job performance, while automation AI use is negatively related to these two outcomes. Furthermore, we focused on the impacts of the ambidextrous use of AI on these two outcomes. The results highlight the positive impacts of complementary use on both outcomes and the negative impact of balance on innovative job performance. TPSC enhances the positive impacts of complementary use on productivity, whereas AITU inhibits the negative impacts of automation and balanced use on innovative job performance.

In the age of AI, organizations face greater trade-offs between performance and technology management. This study contributes to the OM literature from the perspectives of operational performance and technology management in three ways. First, it distinguishes among different AI implementations and their diverse impacts on productivity and innovative performance. Second, it identifies the different conditions under which automation AI use and augmentation are superior. Third, it extends the ambidextrous perspective by becoming an early adopter of this approach to explore the implications of different types of AI use in light of contingency factors.

Create and share a MATLAB library that performs data augmentation algorithms for audio data. This study aims to help machine learning researchers to improve their models using the algorithms proposed by the authors.

The authors structured our library into methods to augment raw audio data and spectrograms. In the paper, the authors describe the structure of the library and give a brief explanation of how every function works. The authors then perform experiments to show that the library is effective.

The authors prove that the library is efficient using a competitive dataset. The authors try multiple data augmentation approaches proposed by them and show that they improve the performance.

A MATLAB library specifically designed for data augmentation was not available before. The authors are the first to provide an efficient and parallel implementation of a large number of algorithms.

The purpose of this paper is to examine the influence of experiential augmentation on product evaluation by consumers. An important distinction is made between product-related experiential augmentation and experiential augmentation of the environment. Furthermore, the research examines how brand familiarity moderates the effect of experiential augmentation.

In two experiments (N = 210 and N = 70), both product-related and environmental experiential augmentation were varied. Participants tasted and evaluated a new coffee product from either a well-known or a fictitious brand.

The findings of the first experiment indicate that product-related experiential augmentation contributes positively to product evaluation for both an unfamiliar and a familiar brand. Experiential augmentation of the environment influences product evaluation negatively, but only in the absence of product-related experiential augmentation. The second experiment tests some possible explanations for this negative effect and shows that it occurs only in the case of a familiar brand.

The findings offer implications for marketing managers seeking to positively influence consumer product evaluations through experiential augmentation. First, marketing managers are advised to make a distinction between product-related experiential augmentation and experiential augmentation of the evaluation environment, and, second, they should take brand familiarity into account when employing experiential augmentation of the environment.

This research contributes to the literature by showing that product-related experiential augmentation and experiential augmentation of the environment differ in the impact they have on product evaluation and providing insight into the relationship between brand familiarity and experiential augmentation.

In recent years, the convolutional neural network (CNN) based deep learning approach has succeeded in data-mining the relationship between microstructures and macroscopic properties of materials. However, such CNN models usually rely heavily on a large set of labeled images to ensure the accuracy and generalization ability of the predictive models. Unfortunately, in many fields, acquiring image data is expensive and inconvenient. This study aims to propose a data augmentation technique to enhance the performance of the CNN models for linking microstructural images to the macroscopic properties of composites.

Microstructures of composites are synthesized using discrete element simulations and Potts kinetic Monte Carlo simulations. Macroscopic properties such as the elastic modulus, Poisson's ratio, shear modulus, coefficient of thermal expansion, and triple-phase boundary length density are extracted on representative volume elements. The CNN model is trained using the 3D microstructural images as inputs and corresponding macroscopic properties as the labels. The comparison of the predictive performance of the CNN models with and without data augmentation treatment are compared.

The comparison between the prediction performance of CNN models with and without data augmentation showed that the former reduced the weighted mean absolute percentage error (WMAPE) for the prediction from 5.1627% to 1.7014%. This significant reduction signifies that the proposed data augmentation method can effectively enhance the generalization ability and robustness of CNN models.

This study demonstrates that data augmentation is beneficial for solving the problems of model overfitting, data scarcity, and sample imbalance for CNN-based deep learning tasks at a low cost. By developing more and advanced data augmentation techniques, deep learning accelerated homogenization will boost the multi-scale computational mechanics and materials.

The numerical solution of contact problems via the penalty method yields approximate satisfaction of contact constraints. The solution can be improved using augmentation schemes. However their efficiency is strongly dependent on the value of the penalty parameter and usually results in a poor rate of convergence to the exact solution. In this paper we propose a new method to perform the augmentations. It is based on estimated values of the augmented Lagrangians. At each augmentation the converged state is used to extract some data. Such information updates a database used for the Lagrangian estimation. The prediction is primarily based on the evolution of the constraint violation with respect to the evolution of the contact forces. The proposed method is characterised by a noticeable efficiency in detecting nearly exact contact forces, and by superlinear convergence for the subsequent minimisation of the residual of constraints. Remarkably, the method is relatively insensitive to the penalty parameter. This allows a solution which fulfils the constraints very rapidly, even when using penalty values close to zero.

A core challenge in background subtraction (BGS) is handling videos with sudden illumination changes in consecutive frames. In our pilot study published in, Sakkos:SKIMA 2019, we tackle the problem from a data point-of-view using data augmentation. Our method performs data augmentation that not only creates endless data on the fly but also features semantic transformations of illumination which enhance the generalisation of the model.

In our pilot study published in SKIMA 2019, the proposed framework successfully simulates flashes and shadows by applying the Euclidean distance transform over a binary mask generated randomly. In this paper, we further enhance the data augmentation framework by proposing new variations in image appearance both locally and globally.

Experimental results demonstrate the contribution of the synthetics in the ability of the models to perform BGS even when significant illumination changes take place.

Such data augmentation allows us to effectively train an illumination-invariant deep learning model for BGS. We further propose a post-processing method that removes noise from the output binary map of segmentation, resulting in a cleaner, more accurate segmentation map that can generalise to multiple scenes of different conditions. We show that it is possible to train deep learning models even with very limited training samples. The source code of the project is made publicly available at https://github.com/dksakkos/illumination_augmentation

The purpose of this research study was to understand the simultaneous competitive and social gains of machine learning (ML) and artificial intelligence (AI) usage in organizations. There was a knowledge hiatus regarding the contribution of the deployment of ML and AI technologies and their effects on organizations and society.

This study was grounded on the dynamic capabilities (DC) and ML and AI automation-augmentation paradox literature. This research study examined these theoretical perspectives using the response of 239 Indian organizational chief technology officers (CTOs). Partial least square-structural equation modeling (PLS-SEM) path modeling was applied for data analysis.

The results indicated that ML and AI technologies organizational usage positively influenced DC initiatives. The findings depicted that DC fully mediated ML and AI-based technologies' effects on firm performance and social performance.

This study contributed to theoretical discourse regarding the tension between organizational and social outcomes of ML and AI technologies. The study extended the role of DC as a vital strategy in achieving social benefits from ML and AI use. Furthermore, the theoretical tension of the automation-augmentation paradox was explored.

Organizations deploying ML and AI technologies could apply this study's insights to comprehend the organizational routines to pursue simultaneous competitive benefits and social gains. Furthermore, chief technology executives of organizations could devise how ML and AI technologies usage from a DC perspective could help settle the tension of the automation-augmentation paradox.

Increased ML and AI technologies usage in organizations enhanced DC. They could lead to positive social benefits such as new job creation, increased compensation to skilled employees and greater gender participation in employment. These insights could be derived based on this research study.

This study was among the first few empirical investigations to provide theoretical and practical insights regarding the organizational and societal benefits of ML and AI usage in organizations because of their DC. This study was also one of the first empirical investigations that addressed the automation-augmentation paradox at the enterprise level.

This research study aims to inquire into the technostress phenomenon at an organizational level from machine learning (ML) and artificial intelligence (AI) deployment. The authors investigated the role of ML and AI automation-augmentation paradox and the socio-technical systems as coping mechanisms for technostress management amongst managers.

The authors applied an exploratory qualitative method and conducted in-depth interviews based on a semi-structured interview questionnaire. Data were collected from 26 subject matter experts. The data transcripts were analyzed using thematic content analysis.

The study results indicated that role ambiguity, job insecurity and the technology environment contributed to technostress because of ML and AI technologies deployment. Complexity, uncertainty, reliability and usefulness were primary technology environment-related stress. The novel integration of ML and AI automation-augmentation interdependence, along with socio-technical systems, could be effectively used for technostress management at the organizational level.

This research study contributed to theoretical discourse regarding the technostress in organizations because of increased ML and AI technologies deployment. This study identified the main techno stressors and contributed critical and novel insights regarding the theorization of coping mechanisms for technostress management in organizations from ML and AI deployment.

The phenomenon of technostress because of ML and AI technologies could have restricting effects on organizational performance. Executives could follow the simultaneous deployment of ML and AI technologies-based automation-augmentation strategy along with socio-technical measures to cope with technostress. Managers could support the technical up-skilling of employees, the realization of ML and AI value, the implementation of technology-driven change management and strategic planning of ML and AI technologies deployment.

This research study was among the first few studies providing critical insights regarding the technostress at the organizational level because of ML and AI deployment. This research study integrated the novel theoretical paradigm of ML and AI automation-augmentation paradox and the socio-technical systems as coping mechanisms for technostress management.

The purpose of this study is to focus on the feeling associated with being rejected by the preferred service brand, and its effect on consumer assessment of the alternate brand.

The data were collected using a self‐administered questionnaire in the context of higher education services targeted at the international market.

Consumers who do manage to get their preferred service brand tend to be more satisfied with the features of the obtained brand and exhibit higher levels of brand loyalty towards that brand. In comparison, consumers who end up with a service brand that is not their first choice seem to have lower levels of satisfaction with and loyalty towards the obtained brand.

A key limitation of this study is the sampling frame. Future studies should replicate this study in different service and product contexts and with different target population. In addition, the disconfirmation of expectations or desires within the framework of preferred brand attainment should be explored.

Managers should ensure that one's service brand is high in the consideration set. This has implications for service branding and brand positioning as well as fulfilling service brand promise through services elements. It also has implications pertaining to winning over new customers and retaining through superior service delivery – particularly the service augmentation elements, and the selection and training of service delivery personnel.

This study provides answers to a crucial question – “Can the number two brand ever achieve a prominent position or is it doomed to remain in the second place waiting to be picked only when consumers do not get their first choice?”

The purpose of this paper is to gain a clear understanding of the impact of uncommon use of knowledge (adaptation and augmentation) on the performance of information systems (IS) departments, and to explore the effects of human-resources management (HRM) practices on uncommon use of knowledge.

A questionnaire-based survey was used to measure the constructs of the research model. A survey package was delivered to project managers or team leads and 133 responses were returned.

The empirical results indicate that knowledge adaptation has a significant effect on departmental performance, whereas knowledge augmentation is more important to innovation than to routine departmental performance. The results also show that, while knowledge adaptation can be enhanced by communication and an uncertainty-avoidance culture, knowledge augmentation is an outcome of shared decision-making, the use of teams, and innovation-based policies.

Given the positive impact of uncommon use of knowledge on IS department performance, future research should explore other factors besides HRM practices to boost it.

The results can serve as guidance for managers looking to select HRM practices to promote uncommon use of knowledge.

This study introduces knowledge adaptation and knowledge augmentation as the component processes of uncommon use of knowledge to the IS discipline, and empirically validates the antecedents and consequences of uncommon use of knowledge using survey data.