Survival Mixture Density Networks
This work addresses computational inefficiency in continuous time survival models for clinical decision-making, offering a faster alternative to neural ODEs while avoiding binning issues in discrete models.
The paper tackles the slow training of neural ODEs in survival analysis by proposing Survival Mixture Density Networks, which use an invertible positive function to map MDN outputs to time-domain distributions, achieving similar or better performance than baselines on metrics like concordance and integrated Brier score across four datasets.
Survival analysis, the art of time-to-event modeling, plays an important role in clinical treatment decisions. Recently, continuous time models built from neural ODEs have been proposed for survival analysis. However, the training of neural ODEs is slow due to the high computational complexity of neural ODE solvers. Here, we propose an efficient alternative for flexible continuous time models, called Survival Mixture Density Networks (Survival MDNs). Survival MDN applies an invertible positive function to the output of Mixture Density Networks (MDNs). While MDNs produce flexible real-valued distributions, the invertible positive function maps the model into the time-domain while preserving a tractable density. Using four datasets, we show that Survival MDN performs better than, or similarly to continuous and discrete time baselines on concordance, integrated Brier score and integrated binomial log-likelihood. Meanwhile, Survival MDNs are also faster than ODE-based models and circumvent binning issues in discrete models.