Deep Learning and Optimization Approaches in E-commerce Systems for Sale Prediction Using Triple Pseudo-Siamese Network with Giant Trevally Optimizer: A Review

Abstract

The rapid growth of e-commerce platforms has produced vast volumes of transactional and behavioral data, enabling advanced predictive analytics for sales forecasting and recommendation systems. Accurate sales prediction is essential for effective inventory management, demand planning, and strategic decision-making in online retail. Traditional statistical models often fail to capture nonlinear patterns and complex relationships in high-dimensional datasets, leading to reduced forecasting accuracy. To overcome these limitations, deep learning techniques have been widely adopted. Among these, Siamese and pseudo-Siamese neural networks are particularly effective in learning similarity relationships across diverse feature spaces.

The Triple Pseudo-Siamese Network extends conventional architectures by integrating multiple input streams to model relationships among customer behavior, product features, and temporal sales data. This enhances the model’s ability to generate accurate predictions in complex e-commerce environments. However, optimizing deep neural networks requires robust algorithms capable of avoiding local optima. The Giant Trevally Optimizer (GTO), a nature-inspired metaheuristic, improves parameter tuning and model convergence.

This review explores recent advancements in deep learning and optimization techniques for e-commerce sales prediction, highlighting hybrid frameworks and identifying research gaps for scalable, intelligent forecasting systems.