Explainable AI-Based Inventory Demand Forecasting for Intelligent Inventory Optimization

Effective inventory management depends heavily on accurate demand forecasting and informed stock control decisions.Tradi-tional forecasting techniques often fail to capture complex demand dynamics influenced by seasonality, promotions, holidays, and external factors.Although machine learning and deep learning models improve predictive accuracy, their black-box nature limits interpretability and reduces managerial trust.This paper proposes an Explainable Artificial Intelligence (XAI)-based framework for inventory demand forecasting and intelligent inventory optimization.The system integrates ARIMA, Prophet, XGBoost, and Long Short-Term Memory (LSTM) models to predict demand patterns.Forecast outputs are used to compute reorder point, safety stock, and economic order quantity while considering service-level targets and cost constraints.To enhance transparency, SHAP and LIME techniques are employed to identify influential demand drivers and generate human-interpretable explanations.A Human-in-the-Loop (HITL) mechanism enables inventory managers to validate and refine system recommendations.Experimental results demonstrate improved forecasting accuracy, reduced stockouts, and lower inventory holding costs compared to traditional methods.