Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{207260,
        author = {Anshul and Tanya Singh and Kushagra Sharma and Deepak Kumar Pathak},
        title = {Sales Forecasting Using Machine Learning Algorithms: A Comparative Study Of Decision Tree, Random Forest, And Extra Tree},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {no},
        pages = {147-152},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=207260},
        abstract = {Sales forecasting is a critical function in business analytics because it supports inventory planning, procurement, budgeting, workforce allocation, and marketing decisions. In retail and food-and-beverage environments, sales are influenced by seasonality, trend, calendar effects, and short-term fluctuations, making simple forecasting rules insufficient for many practical settings. 
This study presents a comparative machine learning framework for forecasting monthly sales using three tree-based regression algorithms: Decision Tree, Random Forest, and Extra Trees. 
The study follows the workflow indicated in the supplied draft, preserving its original focus on historical sales prediction through date-derived features, lag variables, data cleaning, and supervised model evaluation. The paper extends the draft into a research-style article by grounding the methodology in standard forecasting and machine learning literature. It argues that while a single decision tree is easy to interpret, ensemble models are better suited to nonlinear sales behavior because they reduce instability and improve generalization. Consistent with the direction reported in the supplied study, Extra Trees emerges as the strongest overall performer, while Random Forest remains highly competitive and robust. 
The findings show that machine learning can provide practical value for sales forecasting when supported by careful preprocessing, feature engineering, and appropriate evaluation. The paper concludes by identifying future improvements such as richer external features, time-aware backtesting, and broader deployment across product categories.},
        keywords = {Sales forecasting, machine learning, Decision Tree, Random Forest, Extra Trees, feature engineering, ensemble learning, food and beverage analytics, predictive modeling.},
        month = {July},
        }