Demand Prediction and Apparel Production Management Using AI-Based Decision Tree

The apparel industry faces significant challenges in demand forecasting due to market volatility, rapid changes in fashion trends, and diverse consumer behavior, especially within e-commerce environments. Traditional forecasting methods such as linear regression and time series models often fall short in addressing the complex dynamics of the modern fashion market. This study presents a novel integration of demand forecasting and size recommendation into a unified AI-based system utilizing the Decision Tree algorithm. The system is designed to predict product demand while also providing personalized clothing size recommendations based on user attributes such as body measurements, style preferences, and seasonal trends. The system was developed using a structured data processing and predictive modeling approach, incorporating user profiles and trend sentiment derived from social media. The evaluation results show that the system achieved an accuracy rate of 87.5% in demand forecasting and 84% user satisfaction for size recommendations. It demonstrated better adaptability and performance compared to traditional methods such as ARIMA. A functional prototype was implemented, allowing users to interactively input data and receive real-time predictions. This study confirms the potential of Decision Tree-based AI models to enhance the shopping experience, reduce product return rates, and optimize inventory management. Future improvements may involve integrating real-time data and advanced technologies such as 3D body scanning to further increase prediction accuracy and personalization in digital fashion retail.