Survival Analysis via Density Estimation

This paper introduces an algorithm that reinterprets survival analysis through the lens of density estimation, addressing the challenge of censored inputs inherent to survival data. Recognizing that many survival analysis methodologies are extensions of foundational density estimation models, our approach leverages this intrinsic relationship. By conceptualizing survival analysis as a form of density estimation, our algorithm postprocesses the density estimation outputs to derive survival functions. This framework allows for the application of any density estimation model to effectively estimate survival functions, thereby broadening the toolkit available for survival analysis and enhancing the flexibility and applicability of existing density estimation techniques in this domain. The proposed algorithm not only bridges the methodological gap between density estimation and survival analysis but also offers a versatile and robust approach for handling censored survival data.