Leveraging AI-Powered Business Intelligence for Data-Driven Decision Making in the Retail Industry

Abstract

The constant innovations in artificial intelligence which have acted as a deciding factor in the fast-paced and competitive retail environment, machine learning (ML) and deep learning (DL) applications are gaining relevance in the industry. The demand for and the scarcity of data analytics experts, however, remain. So, there is an urgent need for more efficient and accessible intelligent forecasting systems to meet the demand. To overcome this problem, this work suggests an Ensemble Random Forest (RF) and Gated Recurrent Unit (Ensemble RF+GRU) model for accurate retail sales forecasting and data-driven decision making. The proposed model is a fusion of the advantage of deriving features from a RF with the advantage of sequential learning of a GRU, which improves the accuracy (ACC) of forecasting and the efficiency of prediction. The Retail Sales Forecasting Dataset provided by Kaggle was used for experimentation and data was preprocessed to enhance the quality. The results of experiments showed that the proposed Ensemble RF+GRU model had an R² value of 94.5%, MSE of 79564.47, RMSE of 282.07, and MAE of 134.62, which were better than the other models such as LR, SVM, RNN, and ANN. The results proposed that the Ensemble model yields correct, reliable, and efficient forecasting performance and is suitable for intelligent decision-making in the retail industry