2024

Permanent URI for this collection

https://escholar.umt.edu.pk/handle/123456789/7668

Browse

Browsing 2024 by Author "Muhammad Awais"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Retail Sales Forecasting Using Machine Learning: A Comparative Analysis of Random Forest, Linear Regression and XGBoost Approaches
    (UMT, Lahore, 2024) Muhammad Awais
    Sales forecasting is critical in the retail industry since it influences strategic planning and decision-making processes. This thesis investigates how machine learning approaches can improve the accuracy of sales estimates in Pakistan's retail sector. The study employs a comprehensive dataset that includes sales data from 36 stores in three regions—North, South, and Central—from October 2022 to October 2023. Products are classified as summer, winter, or regular. The study looks into how many Machines learning models, such as Extreme Gradient Boosting (XGBoost), Linear Regression, and Random Forest Regression, can effectively forecast sales trends. Data visualization techniques such as box plots, bar charts, and correlation heatmaps are used to understand product category and area sales patterns. These models' performance is evaluated using measures such as 𝑅𝑅 2 and 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅. The data demonstrate considerable seasonal and geographical fluctuations in sales, providing useful information for inventory management and marketing strategies. This study shows how advanced machine learning algorithms may improve prediction accuracy, allowing businesses to optimize operations and avoid costs due to overstock or stockouts. The study also identifies limits and proposes directions for future research, such as incorporating new data sources and investigating more advanced models such as 𝑅𝑅𝑅𝑅𝑅𝑅 and 𝑇𝑇𝑇𝑇𝑇𝑇.