Artificial Intelligence Techniques for E-Commerce System for Sale Prediction Using Triple Pseudo-Siamese Network with Giant Trevally Optimizer: Trends and Challenges

Abstract

The rapid expansion of e-commerce platforms has produced massive volumes of digital data, creating significant opportunities for advanced predictive analytics systems capable of accurately forecasting product demand. Sales prediction plays a crucial role in e-commerce operations as it enables businesses to optimize inventory management, improve supply chain planning, and enhance overall customer satisfaction. Traditional statistical forecasting methods such as regression and ARIMA models have been widely applied in retail prediction tasks; however, they often fail to capture nonlinear relationships and complex patterns inherent in large-scale e-commerce datasets. Consequently, artificial intelligence and deep learning approaches have become increasingly important for modeling sales trends and consumer demand behavior. This review examines a range of AI-based techniques for e-commerce sales prediction, with particular focus on advanced deep learning architectures and optimization strategies used to improve forecasting performance. It further analyzes contributions in machine learning algorithms, neural networks, and predictive analytics frameworks applied in digital commerce environments. Additionally, the study compares existing forecasting models and identifies key limitations in current approaches. Finally, it highlights emerging challenges and potential future research directions in developing more accurate and efficient intelligent forecasting systems for online retail ecosystems.