USING RISK FACTORS FOR DISEASE TO PREDICT PROBABILITY OF CONTRACTING A DISEASE IN A MACHINE LEARNING-BASED PRODUCT THAT RECOMMENDS INTERVENTIONS TO INCREASE HEALTH AND LONGEVITY

Abstract

In previous papers, we have described a methodology of using meta-regression and Bayesian statistics to create a machine learning model that combines scientific research on wellness and longevity and responses from people’s lifestyle questionnaires to make recommendations on what people can do to live longer and healthier. One of the goals of this software is to use data collected from logs of people’s lifestyle choices to update the model and increasingly improve and personalize the recommendations made. One challenge faced here is that the elapsed time between lifestyle choices and the onset of disease may take years, making it difficult to make