The integration of machine learning (ML) techniques into software engineering has revolutionized the field, offering novel solutions to long-standing problems and enabling the creation of more sophisticated, efficient, and reliable software systems. This paper explores the advances in machine learning and their impact on software engineering, focusing on key ML algorithms, foundational theories, and the emerging role of Graph Neural Networks (GNN). Through a comprehensive literature review, we highlight the significant contributions and applications of ML in software engineering. The paper details the use of prominent software libraries and frameworks, such as Scikit-learn, TensorFlow, and Stable-Baselines3, discussing their features, implementation details, and performance benchmarks. We also examine the challenges faced in ML applications, including data quality, preprocessing, and the development of hybrid models. The discussion extends to the future directions of ML in real-world applications, emphasizing its potential in cybersecurity, healthcare, smart cities, and the Internet of Things (IoT). Our findings underscore the transformative potential of ML in software engineering and provide a roadmap for future research and practical applications in this dynamic field.