Year: 2024
Type: conference-paper
Abstract: The article considers the current features of the logistics component and key current trends in the field of logistics. Particular attention is paid to the analysis of problems that are successfully solved with the help of machine optimization of logistics routes, accurate demand forecasting, efficient management of warehouse inventory management, predictive maintenance, product quality control, optimization of last-mile delivery optimization, risk management, and environmental sustainability. In this paper analyzes the most common and effective approaches used in machine learning, including artificial neural networks and decision trees. The advantages of using artificial neural networks in demand forecasting are described, the essence of the of the decision tree methodology, and the advantages of their application, in particular, the possibility of making of making informed management decisions, risk reduction, support for strategic planning, simplicity and simplicity and clarity of use, as well as the non-parametric nature of this method, which allows avoiding mistakes, associated with incorrect assumptions about the distribution of data. The practical application of artificial neural networks and decision trees in business is analyzed separately, which made it possible to demonstrate their effectiveness in solving real problems in the modern logistics system. The article describes the CCD methodology, which covers various key stages, such as data collection, cleaning and data processing, extraction of useful information, and interpretation of the identified patterns. Important attention is paid to the results that demonstrate a significant improvement in forecasting accuracy, as well as a significant reduction in the corresponding errors. In addition, a decision tree-based regression model is considered, which, due to its ability to effectively work with nonlinear data, allows to increase the overall productivity of an enterprise by optimizing decision-making processes and improving the efficiency of of resource management.
Authors: Maksym Alieksieiev, Mykola Nebava
Institution: Vinnytsia National Technical University
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: gold
OpenAlex ID: W4405876701