Machine Learning: Modern Techniques and Mathematical Approach to Neural Networks

Abstract

This thesis explores the mathematical foundations of machine learning algorithms, providing a comprehensive exploration of various key techniques and models. It covers fundamental concepts and methodologies in regression, classification, support vector machines (SVM), decision trees, neural networks, and perceptrons. Each of these algorithms is analyzed in terms of their mathematical underpinnings, operational mechanisms, and practical applications. The study aims to elucidate the core principles that drive these algorithms, offering insights into their theoretical and practical aspects. By understanding the mathematics behind these models, this research contributes to a deeper appreciation and effective utilization of machine learning techniques in solving complex real-world problems.