Leveraging Machine Learning for Superior E-Commerce Performance: Advanced Sales Forecasting Techniques for Business Growth

Dhaval Gogri, Niketu Dave, Ashish Korpe

Abstract: In today's highly consumer-oriented business environment, any enterprise in search of high sales results must find an optimal relationship between supply demand and costs of storage. Adequate sales forecasts are especially important when it comes to helping firms enhance their operations. The advancement in Artificial Intelligence has resulted in an increase in the methods used in the solution of the forecasting problem. Concerning essential algorithms for sales forecasting, this paper employs RF, XGBoost, ARIMA, and KNN. These algorithms were applied to a Walmart real-world sales data set, which is illustrated in Figure 4. The performance of the algorithm results was further compared with each to make use of the R2 score, MAE, and RMSE. The results depicted below show that the RF and XGBoost algorithms offered higher accuracy compared to the other algorithms. The results recommended that the RF model presents the best performance with R?=96% and MAE=19.85. Furthermore, the XGB model presents a nearly comparable performance with an R? score of 94% and an MAE of 55.34. In contrast, ARIMA and KNN models demonstrate comparatively lower accuracy. As a result, the study could be valuable in illustrating how Assuming, the appropriate utilization of the results of machine learning approaches could enhance the actuality of sales forecasts aimed at e-business realization. Future work may thus look at how more data could be incorporated, including customer satisfaction survey results and current market trends, for the purpose of improving the credibility of the model.

Keywords: Sales forecasting, machine learning, e-commerce, sentiment analysis, business sector, retail

How to Cite?: Dhaval Gogri, Niketu Dave, Ashish Korpe, "Leveraging Machine Learning for Superior E-Commerce Performance: Advanced Sales Forecasting Techniques for Business Growth", Volume 13 Issue 10, October 2024, International Journal of Science and Research (IJSR), Pages: 1044-1051, https://www.ijsr.net/getabstract.php?paperid=SR241014114523, DOI: https://dx.doi.org/10.21275/SR241014114523