Risk management of investment projects in flexible environments

Шульга, Ольга Антонівна (2026) Risk management of investment projects in flexible environments Підприємництво та інновації (41). pp. 161-165. ISSN 2707-6237

Text Shulha_OA_PI_41_2026.pdf Download (292kB)

Abstract

The relevance of the topic is determined by the high level of uncertainty of modern markets, where traditional static risk management methods are ineffective due to the inability to take into account dynamic changes in the development process. The purpose of the study is to develop a comprehensive methodology for predictive risk management of investment projects based on the integration of hybrid calendar planning models and machine learning algorithms for advanced forecasting of financial costs. The methodological basis of the study is the methods of system and comparative analysis for the study of Agile, Scrum and Waterfall models, as well as the concept of synergy of strategic and operational management. To build an integrated analytical circuit, methods of mathematical transformation of operational metrics of the Kanban system into time parameters of the Gantt chart were used. The process of modeling financial deviations is based on the application of ensemble methods of machine learning, in particular, random forest and gradient boosting algorithms, which allow for effective processing of nonlinear relationships and structural noise in input arrays of technical and economic data. As a result of the study, a two-level concept of investment project management was proposed, where the Gantt chart provides strategic control of milestones, and the Kanban system provides operational flexibility. A logical-mathematical mechanism for extrapolating the metrics Lead Time, Cycle Time, and Throughput onto the timeline of a linear graph was developed, which minimizes the subjectivity of expert assessments. A classification of specific time and resource risks at the interface of methodologies was formed. The practical value of the results obtained lies in creating a theoretical basis for building automated systems for intelligent decision-making support in investment management. The use of a predictive ML module provides high accuracy in determining the final cost of works, which allows investors to dynamically redistribute capital, significantly optimize the size of project reserve funds, avoid excessive freezing of liquid funds, and direct the released resources to parallel investment areas.

Item Type:	Article
Uncontrolled Keywords:	investment projects; risk management; flexible environments; Gantt chart; Kanban; predictive analysis; machine learning; cost forecasting; artificial intelligence; project implementation time planning; project budget
Subjects:	Статті у базах даних > Index Copernicus Статті у періодичних виданнях > Фахові (входять до переліку фахових, затверджений МОН)
Divisions:	Факультет економіки та управління > Кафедра управління
Depositing User:	Ольга Антонівна Шульга
Date Deposited:	07 Sep 2026 16:14
Last Modified:	07 Sep 2026 16:14
URI:	https://elibrary.kubg.edu.ua/id/eprint/59477

Actions (login required)

View Item