Forecasting the Development Trends of the IT Market Using Machine Learning Methods
Open Access
- 30 November 2025
- journal article
- Published by Lviv Polytechnic National University in Computer systems and network
- Vol. 18 (2) , 163-183
Abstract
The article explores approaches to forecasting the development trends of the IT market using machine learning methods. The relevance of the research is driven by the high dynamics of the digital economy, rapid technological changes, and the need for scientifically grounded analytical tools in the IT domain. The purpose of the study is to develop a forecasting model capable of identifying patterns in socio-economic, technological, and behavioral indicators that determine the state and prospects of IT market development. The study employs time series analysis, correlation-regression modeling, and machine learning techniques − specifically, the Random Forest, Gradient Boosting, and Long Short-Term Memory algorithms. The dataset integrates macroeconomic indicators (GDP, exchange rate, consumer price index), IT market metrics (average salary, job demand, IT service export volume), and behavioral data (Google Trends search indices). Data preprocessing, feature normalization, and stationarity testing of time series were performed. The results show that the Random Forest and LSTM models achieve the highest forecasting accuracy (R² > 0.85), effectively capturing both short- and medium-term trends in IT market development. Key predictors of market dynamics were identified, including the level of consumer spending, average salary, and IT query popularity index. The practical significance of the research lies in the possibility of applying the proposed models to support strategic decision-making in the digital economy, educational and workforce planning, and forecasting technological trends in the national IT sector.Keywords
This publication has 5 references indexed in Scilit:
- Machine learning, stock market forecasting, and market efficiency: a comparative studyInternational Journal of Data Science and Analytics, 2025
- Gradient Boosting-Based Simultaneous Classification and Regression ApproachIEEE Access, 2025
- TRENDS IN ENSEMBLE LEARNING AND MODEL OPTIMIZATION2025
- Comparative Analysis of ARIMA and LSTM Models for Stock Price Prediction2024
- Recurrent Neural Networks for Time Series Forecasting: Current status and future directionsInternational Journal of Forecasting, 2021