METHODS OF DATA MINING USING MACHINE LEARNING

Abstract

Relevance . In the context of the continuous growth of information volumes across various fields of the economy, science, and technology, the problem of effective processing and analysis of large-scale data has become increasingly urgent. Traditional analytical methods are no longer capable of providing fast and accurate extraction of useful knowledge and patterns from massive information flows. The response to this challenge lies in the methods of data mining, which are based on modern machine learning technologies. These methods enable the automatic discovery of hidden patterns, the generation of accurate predictions, and the support of data-driven decision-making in real time. Given the rapid development of digitalization, artificial intelligence, and the need for prompt decision-making in a competitive environment, the relevance of developing and improving data mining methods is growing steadily. The object of research is process of data mining using machine learning methods, namely the set of algorithms, models, tools, and approaches that ensure the detection of hidden patterns, anomalies, and structures in large volumes of heterogeneous information. Purpose of the article. This study explores contemporary approaches to intelligent data analysis based on machine learning techniques and assesses their effectiveness across a range of application domains. The article aims to provide a structured overview of state-of-the-art algorithms and to evaluate their respective advantages and limitations in processing large-scale and highdimensional datasets. Research results. A systematic analysis of key data mining methods based on machine learning algorithms was carried out. It was found that the most effective approaches for processing large and heterogeneous datasets include classification, clustering, regression analysis, and dimensionality reduction techniques. Deep neural networks demonstrated effectiveness when applied to unstructured data such as text, images, and time series. The study revealed that the appropriate choice of algorithm depends not only on the data type but also on the specific nature of the task. A comparative assessment of tools showed that the Python ecosystem offers the greatest flexibility, while AutoML platforms simplify model deployment for users with limited programming experience. The research also included a review of recent publications that confirm the practical value of machine learning in real-world use cases. Overall, the findings indicate that machine learning is a driving force behind the evolution of data mining methods, enabling accurate, scalable, and adaptive data processing in the context of modern digital transformation. Conclusions. Machine learning has significantly expanded the capabilities of intelligent data analysis by enabling the automatic detection of patterns, forecasting, and decisionmaking based on large volumes of information. The study demonstrates the effectiveness of various algorithms in tasks such as classification, clustering, regression, and deep learning. Python-based tools and cloud platforms have been identified as the most convenient environments for implementing analytical models. A promising direction lies in the development of explainable AI and hybrid approaches that combine algorithmic precision with domain-specific expertise.