The Intelligent Decision Support System in Comprehensive Management of Enterprise Assortment and Inventory Under Rapid Market Transformations

Abstract

The article explores the theoretical and practical aspects of managing enterprise assortment and inventory under fast-changing market conditions, which are characterized by high demand dynamics, external uncertainties, and growing data volumes. The feasibility of using intelligent decision support systems as a tool to improve the efficiency of assortment and inventory management is substantiated. A conceptual model of such a system is proposed, representing an integrated information-analytical environment that ensures a full data processing cycle – from collection to managerial decision-making. The system’s structure includes a number of interconnected functional modules. The data collection and integration module ensures the consolidation of information from internal and external sources, as well as its cleaning, normalization, aggregation, and structuring. The data storage and management module is responsible for organizing a unified information space, controlling data quality, keeping it up to date, indexing, and ensuring quick access to information. The analytics module implements a range of statistical analysis methods, including correlation, regression, and cluster analysis, as well as time series decomposition, which helps identify patterns and factors affecting demand. The forecasting subsystem is based on a combination of classic time series models and modern machine learning methods, including LSTM neural networks and ensemble approaches, which improves forecasting accuracy in conditions of unstable demand. The optimization and decision support module implements economic and mathematical models for inventory and assortment management, as well as multi-criteria optimization problems. An important feature is scenario modeling, which allows evaluating different market development options and choosing the most resilient managerial decisions considering risks and uncertainty. The user interface provides visualization of analysis results through dashboards, charts, and analytical reports, as well as recommendations for inventory management and assortment planning. The system’s interactivity makes it possible to run «what–if» scenarios and support real-time decision-making. The proposed conceptual model helps improve demand forecasting accuracy, optimize inventory levels, reduce costs, and minimize the risks of shortages or surpluses, which enhances the company’s operational efficiency and strengthens its economic security.