Managing various aspects of the Supply Chain (SC) has become increasingly challenging in today’s complex business landscape. To improve profitability, boost sales, and enhance customer satisfaction, it is crucial to explore future possibilities by adjusting key relational parameters. However, traditional forecasting methods often fail to provide accurate insights and are time-consuming. These limitations can be overcome using Artificial Intelligence (AI) algorithms such as Machine Learning (ML) and Deep Learning (DL). CatBoost algorithm is an ensemble-based ML model that can handle categorical variables effectively in its architecture, whereas other ML and DL models fail and require explicit encoding techniques. In this study, a predictive modeling approach using CatBoost to optimize supply chain processes using a mathematical approach was proposed. CatBoost evaluates the model on an e-commerce dataset through empirical analysis by tuning various hyperparameters to enhance prediction efficiency. A computational time limit of ten minutes was used to ensure practicality. Using regression and classification frameworks, this approach involves predicting sales, profit, and delivery times, and identifying potential customers. Consequently, analyzing the behavior of the learning rate and its impact on the performance metrics indicated that increasing the learning rate can lead to improved model performance.

In today’s dynamic business environment, the accurate prediction of sales orders plays a critical role in optimizing Supply Chain Management (SCM) and enhancing operational efficiency. In a rapidly changing, Fast-Moving Consumer Goods (FMCG) business, it is essential to analyze the sales of the products and accordingly plan the supply. Due to low data volume and complexity, traditional forecasting methods struggle to capture intricate patterns. Domain Adversarial Neural Networks (DANNs) offer a promising solution by integrating transfer learning techniques to improve prediction accuracy across diverse datasets. This study presents a new sales order prediction framework that combines DANN-based feature extraction and various machine learning models. The DANN method generalizes the data, maintaining the data behavior’s originality. The approach addresses challenges like limited data availability and high variability in sales behavior. Using the transfer learning approach, the DANN model is trained on the training data, and this pre-trained DANN model extracts relevant features from unknown products. In contrast, Machine Learning (ML) algorithms are used to build predictive models based on it. The hyperparameter tuning of ensemble models such as Decision Tree (DT) and Random Forest (RF) is also performed. Models like the DT and RF Regressor perform better than Linear Regression and Support Vector Regressor. Notably, even without hyperparameter tuning, the Extreme Gradient Boost (XGBoost) Regressor model outperforms all the other models. This comprehensive analysis highlights the comparative benefits of various models and establishes the superiority of XGBoost in predicting sales orders effectively.