Comparative Analysis of Deterministic vs. ML-Based Stochastic Mathematical Inventory Models

Abstract

This paper compares two main ways of managing inventory the traditional deterministic models and modern machine learning (ML)-based stochastic models. In deterministic models, demand is treated as fixed and known in advance. These models are simple, easy to use, and work well when demand changes very little. However, they often fail when real-life demand is uncertain or changes quickly. The ML-based stochastic models use data-driven methods to predict demand by learning from past trends, prices, seasons, and other factors. Instead of assuming one fixed demand value, they use probability and statistics to show how demand may vary. This helps companies make better decisions about how much to order and when to restock. The comparison shows that ML-based models can adapt quickly, reduce total costs, and maintain better service levels. Deterministic models are still useful for stable markets, but ML-based stochastic models perform better in uncertain, fast-changing environments because they include learning, prediction, and real-time adjustment in inventory control.