Fig. 4: Prevalence by sociodemographic group and sensitivity of the machine learning predictions when using a relative or an absolute cutoff for targeting.
From: Predicting control of cardiovascular disease risk factors in South Asia using machine learning
Red bars indicate grouping by sex (male/female), green bars indicate grouping by age (below age 50/age 50 and older), blue bars indicate grouping by education (above secondary education, secondary education, below secondary education). a Prevalence for not achieving control. b Prevalence for not achieving improvement. c Sensitivity for not achieving control, targeting top 25%. d Sensitivity for not achieving improvement, targeting top 25%. e Sensitivity for not achieving control, absolute targeting. f Sensitivity for not achieving control, absolute targeting.
