Machine Learning for Time-to-Event Prediction and Survival Clustering: A Review from Statistics to Deep Neural Networks

Survival analysis is a statistical method used in computational biology to investigate the time until the occurrence of an event of interest, such as death or disease recurrence. It plays a crucial role in analyzing and understanding time-to-event data in various medical and biological studies. Deep learning, as a subset of artificial intelligence, has shown remarkable success in diverse domains, such as image recognition and natural language processing. Its ability to automatically extract complex patterns and features from high-dimensional data makes it highly promising for enhancing survival analysis. While existing reviews have primarily focused on traditional statistical methods and conventional machine learning approaches in survival analysis, a critical aspect that has been largely overlooked is the integration of deep learning techniques. We not only considered traditional statistical methods and conventional machine learning approaches but also further incorporated deep learning methods. With this review article, we hope to provide researchers, doctors, and biologists with a comprehensive framework for understanding survival analysis and to foster the development of survival prediction and personalized medicine.