Original title: Bayesian inference of a new Mallows model for characterising symptom sequences applied in primary progressive aphasia
Authors: Beatrice Taylor, Cameron Shand, Chris J. D. Hardy, Neil Oxtoby
This study delves into using machine learning to grasp complex datasets, particularly in understanding disease experiences for more equitable healthcare. The focus is on using Bayesian inference to analyze symptom sequences, a task that comes with its challenges. They’ve tailored the Mallows model to handle partial rankings and right-censored data, employing a custom fitting approach called MCMC. Testing it on synthetic and primary progressive aphasia datasets revealed its strength in uncovering average orders and gauging ranking differences, offering potential insights into how symptoms unfold clinically. However, limitations emerged around scaling the model and working with smaller datasets, showing there’s still ground to cover in these areas.
Original article: https://arxiv.org/abs/2311.13411