Machine learning models demonstrate remarkable potential to correct self-reported anthropometric data and improve the accuracy of obesity prevalence estimates.