This paper provides a comprehensive survey of recent research on imbalanced learning, covering its latest developments, applications, and future directions. The authors highlight the challenges of imbalanced class distributions in machine learning and deep learning, emphasizing the need for effective strategies to address the bias towards majority classes. The paper categorizes existing methods into five types: general methods, ensemble learning methods, imbalanced regression and clustering, long-tail learning, and imbalanced data streams. It further explores detailed methods such as data-level approaches, algorithm-level techniques, hybrid methods, and ensemble learning. The survey also discusses real-world applications of imbalanced learning across various domains, including management science, engineering, and healthcare. Additionally, the paper identifies six new research challenges and directions in imbalanced learning, emphasizing the importance of addressing these issues for future advancements. The authors also present a systematic review of recent research methodologies, including statistical classification research and preliminary statistics, to ensure comprehensive coverage of the field. The paper concludes with a discussion of the main contributions, including a unified review of imbalanced learning and deep imbalanced learning, a comprehensive survey of long-tail learning and imbalanced machine learning applications, and the identification of six new research challenges and directions. The survey aims to provide a valuable resource for researchers and practitioners in the field of imbalanced learning, facilitating further exploration and development of effective solutions.This paper provides a comprehensive survey of recent research on imbalanced learning, covering its latest developments, applications, and future directions. The authors highlight the challenges of imbalanced class distributions in machine learning and deep learning, emphasizing the need for effective strategies to address the bias towards majority classes. The paper categorizes existing methods into five types: general methods, ensemble learning methods, imbalanced regression and clustering, long-tail learning, and imbalanced data streams. It further explores detailed methods such as data-level approaches, algorithm-level techniques, hybrid methods, and ensemble learning. The survey also discusses real-world applications of imbalanced learning across various domains, including management science, engineering, and healthcare. Additionally, the paper identifies six new research challenges and directions in imbalanced learning, emphasizing the importance of addressing these issues for future advancements. The authors also present a systematic review of recent research methodologies, including statistical classification research and preliminary statistics, to ensure comprehensive coverage of the field. The paper concludes with a discussion of the main contributions, including a unified review of imbalanced learning and deep imbalanced learning, a comprehensive survey of long-tail learning and imbalanced machine learning applications, and the identification of six new research challenges and directions. The survey aims to provide a valuable resource for researchers and practitioners in the field of imbalanced learning, facilitating further exploration and development of effective solutions.