Economic evaluation of the effectiveness of implementing decision support systems in logistics using machine learning algorithms

Abstract

The relevance of the study stems from the increasing complexity of logistics systems, the deepening of digital transformation of management processes, and the rising uncertainty in demand, supply chain parameters, and external regulatory and market conditions affecting logistics operations.

The purpose of the article is to provide theoretical and methodological substantiation of the economic feasibility of using management decision support systems in logistics based on machine learning algorithms, and to identify directions for increasing their effectiveness in logistics management processes.

Research methods include systemic and structural-functional analysis, generalisation and comparison of scientific approaches to the economic evaluation of intelligent management systems, logical-analytical methods, and conceptual analysis of the costs, outcomes, and risks of operating management decision support systems in logistics processes.

It has been established that the economic effect of applying machine learning algorithms is complex and manifests as reduced logistics costs, increased process stability and predictability, and reduced management risks. The expediency of an integrated approach to economic evaluation, combining cost, performance, and risk parameters, has been proven. The main problems in implementing and evaluating ML systems have been identified: data quality and heterogeneity, limited model adaptability to dynamic changes in the logistics environment, and insufficient interpretability of algorithmic recommendations.

It is substantiated that management decision support systems based on machine learning algorithms are an economically feasible tool for increasing the efficiency of logistics management, provided they are purposefully designed, phased in, and functionally coordinated with management processes of planning, budgeting, and controlling.

Prospects for further research include the development of formalised models for the quantitative assessment of the long-term economic efficiency and sustainability of logistics systems, as well as improvements to methods for the economic analysis of intelligent management decisions.