Data Warehousing and Analytics for E-Commerce Using EfficientDet Model (EDM)

In the rapidly evolving digital marketplace, ecommerce websites generate vast amounts of data that require efficient storage, processing, and analysis to derive actionable business insights. This paper presents a comprehensive project aimed at building a scalable data warehouse and analytics tool for an Amazon e-commerce dataset using Hadoop Hive and Apache Spark. Our solution provides a high-performance data warehousing environment coupled with machine learning-based predictive analytics to forecast sales. By leveraging big data technologies, we provide real-time insights and accurate sales predictions highlighting the potential of our approach to enhance decision making processes for e-commerce business. This Solution is particularly beneficial for optimizing inventory and marketing strategies. Our EfficientDet Model (EDM) demonstrates a prediction accuracy of $85 \%$ on test data, while our Spark-based analytics pipeline processes queries $\mathbf{4 0 \%}$ faster than traditional SQL databases.