METHODOLOGY FOR ANALYZING SOCIO-ECONOMIC DATA OF TERRITORIAL COMMUNITIES

Abstract

In the current era of digital transformation and rapid data growth, territorial communities face challenges that require innovative approaches to analyzing socio-economic information. Socio-economic data of territorial communities is characterized by high heterogeneity, non-stationarity, and dynamism, which significantly complicate its processing using traditional methods. A pressing task is to enable operational decision-making based on the analysis of such data, contributing to improved resource management efficiency and the quality of life of community residents.

This article presents a novel methodology for analyzing socio-economic data of territorial communities, combining classification, clustering, and hybrid analysis with forecasting. The methodology ensures adaptability to non-stationary data, integration of heterogeneous information sources, and real-time processing capability.

A new method for constructing training datasets for classification and forecasting of non-stationary data is proposed, based on RFM analysis and clustering. Additionally, an ensemble method of adaptive predictive models is developed for the first time, ensuring high accuracy in forecasting multidimensional processes, which is critically important for operational decision-making.

The methodology provides a systematic approach to the analysis of socio-economic data, allowing to adapt the analysis process to the specific conditions of the TG, increasing decision-making and optimizing resource management. The research outcomes can be utilized for efficient management of socio-economic processes in territorial communities, adaptive forecasting, and the development of intelligent decision-support systems.

Keywords: territorial