Sales Forecasting Using Machine Learning Deployed on Cloud

This study developed a cloud-based sales forecasting system using machine learning models. Experimental results show that Random Forest Regression performs better than Linear Regression in predicting sales with higher accuracy under changing and non-linear demand conditions. The cloud-based implementation allows users to obtain real-time predictions through a simple web interface. Hence, the proposed system can effectively support business planning and sales decision- making.

[1]     L. Breiman, “Random Forests,” Machine Learning, vol. 45, no. 1, 2001. [2]     R. J. Hyndman and G. Athanasopoulos, Forecasting: Principles and Practice, 3rd ed., Melbourne, Australia: OTexts, 2021. [3]     M. R. Hasan, M. M. Rahman, and S. Islam, “A Compara- tive Study on Forecasting of Retail Sales,” arXiv preprint arXiv:2203.06848, 2022. [4]     E. Zunic, E. Avdovic, and M. Music, “Application of Facebook Prophet Algorithm for Successful Sales Fore- casting,” arXiv preprint arXiv:2005.07575, 2020. [5]     E. Alpaydin, Introduction to Machine Learning, 3rd ed., Cambridge, U.K.: MIT Press, 2014. [6]     K. L. Mench, S. Rossi, and A. Micheli, “Com- paring Prophet and Deep Learning on ARIMA in Forecasting Wholesale Food Prices,” arXiv preprint arXiv:2107.12770, 2021. [7]     R. Laframboise and R. Vaidy, “Application of Machine Learning Techniques for Supply Chain Demand Fore- casting,” European Journal of Operational Research, vol. 184, no. 3, 2008. [8]     S. Makkopoulos, “The M4 Competition: Results, Find- ings, and Conclusions,” International Journal of Fore- casting, vol. 34, no. 4, 2018. [9]     T. M. Choi, S. W. Wallace, and Y. Wang, “Big Data Analytics in Operations Management,” Production and Operations Management, vol. 27, no. 10,2018. [10]    T. Hall and K. Rasheed, “A Survey of Machine Learning Methods for Time Series Prediction,” Applied Sciences, vol. 15, no. 2, 2025.