Determining the current role, modern aspects and future orientations of machine learning algorithms

This thesis comprehensively examines the current role of machine learning, its theoretical foundations, application domains, and future directions. As the need for data-driven decision-making increases, machine learning systems have become indispensable across many sectors. In this context, the study is structured into three main sections, covering both theoretical and practical aspects. The first section focuses on the core components of machine learning, such as data dependency, hardware requirements, feature engineering, execution time, model selection, and interpretability. It provides a comparative evaluation of traditional and deep learning approaches and analyzes supervised, unsupervised, and reinforcement learning in detail. The second section explores the use of machine learning in the field of cybersecurity and includes a practical analysis using the NSL-KDD dataset within the context of intrusion detection systems. By utilizing the KNIME platform, the performance of various algorithms—such as decision trees, random forests, SVM, and Naive Bayes—was compared. The results demonstrate that machine learning is a powerful tool, particularly for anomaly detection and network security. The third section discusses the future trends of machine learning, including a case study on sentiment analysis of Amazon product reviews. It also addresses current challenges such as model interpretability, data quality, scalability, and regulatory compliance, along with proposed solutions to these issues. In conclusion, this thesis presents a holistic view of both the current applications and strategic future significance of machine learning. It illustrates how machine learning can be effectively applied across a wide range of domains, from decision support systems to security, commerce, and healthcare.