SALSTM: An improved LSTM algorithm for predicting the competitiveness of export products

Abstract

As international trade has been developing at an unprecedented rate, export product competitiveness is significant in a country's trade system. At present, neural network algorithms are extensively used in economic forecasting. Scholars have verified the effectiveness of neural networks, especially long short term memory (LSTM) model, compared with other nonlinear prediction methods. However, there are still research blanks in forecasting economic indicators related to export product competitiveness, and there is little research on the improved LSTM model combined with other algorithms. To address this limitation, this article established a quantitative index system for export product competitiveness composed of four highly recognized indicators such as revealed comparative advantage index, net export revealed comparative advantage index, international market share index, and economic value added. After collecting the relevant economic data, the single LSTM model was used to predict the future trend of China's machinery and equipment export product competitiveness. Based on the shortcomings exposed in applying single LSTM prediction, the article innovatively designs a combined forecast model of LSTM and seasonal autoregressive integrated moving average, and whose effectiveness and superiority are proved by the mean square error value. Finally, the article points out that this model plays an important role in national trade policy making. It is optimistic about the development of deep learning in international trade product competitiveness indicators prediction.

Managing artificial intelligence in international business: Toward a research agenda on sustainable production and consumption

“…As an example, managers can deploy AI-enabled computer vision to detect production defects early and achieve real-time production adjustment. Furthermore, managers can equip assembly lines with AI to constantly monitor the production process and autonomously, which aids in reducing operational costs and automating product testing (Yu, Fletcher, & Buck, 2022;Yu, Liu, et al, 2022).…”

A Spatio-Temporal Siamese Neural Network for Multimodal Handwriting Abnormality Screening of Parkinson’s Disease

“…Previously, there were a number of machine learning methods and technologies applied in the feld of data classifcation [4][5][6]. With regard to the challenging PD diagnosis problem, these methods provide superior specifcations.…”

“…Equation (6) shows the calculation process of the Octconv unit. Te input of Octconv consists of high-frequency feature X H and low-frequency feature X L , with the two outputs Y � Y H , Y L 􏼈 􏼉.…”

Time series change detection using reservoir computing networks for remote sensing data

“…Recursive neural networks 4,5 have a feedback connection mechanism and are suitable for dynamic data processing. Recursive connections allow the network to capture temporal features in continuous inputs, and the temporal dependence of the inputs can be embedded in their dynamic behavior.…”

“…The current state depends not only on the current input but also on previous input states; these operations increase the computational cost of training the network. [1][2][3] Recursive neural networks 4,5 have a feedback connection mechanism and are suitable for dynamic data processing. Recursive connections allow the network to capture temporal features in continuous inputs, and the temporal dependence of the inputs can be embedded in their dynamic behavior.…”