Detecting the crisis of supply chain management is an important subject for the sustainable management of e-commerce. The purpose of this study was to explore the application of Q-technique to the detection of supply chain crises in the imported food industry under the influence of COVID-19. Literature analysis, expert consultation, and Q-technique methods were adopted. We used an expert consultation and questionnaire survey combined with Q-technique using the imported food industry in Taiwan as an example. Fifteen business owners were interviewed to identify priorities for supply chain crises, and 137 questionnaires were collected to determine the strength of each item. By investigating suppliers, distributors, wholesalers, platform operators, online sellers, and logistics operators, this study obtained the priority of supply chain crisis items and the strength of factors for each item. The results show that Q-technique can effectively detect the crisis of cross-border supply chain management, clarifying import food industry e-commerce supply chain crisis items and determining their priority and strength, to help enterprises make emergency decisions regarding supply chain operation, and they can also be used as a reference for enterprises to respond to the crisis.
12 April 2023 in International Journal of Information Technology & Decision Making
Due to the powerful feature learning and classification capabilities of convolutional neural networks, they have been widely applied in fields such as image recognition and classification, object detection, and semantic segmentation. Therefore, enhancing the efficiency and accuracy of convolutional neural networks in deep learning has become increasingly important. In convolutional neural networks, the initial design of the convolutional kernels significantly influences the efficiency of deep learning iterations. This paper focuses on the recognition and classification of image tables, proposing a method of customizing the initialization of convolutional kernel parameters based on several local elements of images and the similarity between their pixel distribution features and the initial parameters within the convolutional kernels. The deep learning of image tables through convolutional neural networks is then conducted on this basis, and compared experimentally with traditional initialization methods such as Normal and Xavier. The experimental results show that the parameter initialization method proposed in this paper achieves notably higher classification accuracy in the early stages of neural network training and overall higher classification accuracy for table recognition.