Heart disease is a prevalent condition and the leading cause of death globally. Its assessment is a complex and costly process. To reduce costs and improve diagnostics, machine learning classification has been utilized since the 1990s, supporting prognosis, diagnosis, screening, monitoring, management, and treatment of various heart diseases, resulting in an extensive body of research. Our previous study contributed to this body of knowledge by conducting a comprehensive systematic mapping study on the use of machine learning classification techniques in cardiology. We reviewed 715 selected studies published from 1997 to December 2023. These studies were systematically categorized based on eight criteria: publication year, type of contribution, empirical study design, type of medical data used, machine learning techniques employed, medical tasks addressed, heart pathology assessed, and classification type. This chapter builds on the previous systematic mapping study by emphasizing the quality of the reviewed studies, with a particular focus on the data used. From the initial 715 papers, high-quality studies were selected for in-depth analysis, and additional studies published after the initial review were also included, resulting in a total of 219 studies. Beyond the criteria evaluated in the previous study, this chapter also examines additional aspects such as data processing tasks, optimization techniques, datasets, validation methods, and accuracy metrics. This study provides valuable insights for future research by synthesizing the key factors that contribute to the successful application of machine learning classification in cardiology.