Misclassification Rates in Hypertension Diagnosis due to Measurement Errors
Camila Friedman-Gerlicz, Isaiah Lilly · SIAM Undergraduate Research Online · 2009
Using a mixture of two normal distributions, we estimate the false positive and false negative errors in the diagnosis of hypertension.Parameters in the mixture are estimated by the expectation-maximization (EM) algorithm.It is shown that both errors depend on cutoff points.Repeated measurements reduce both errors dramatically.The number of repeated measurements is recommended through a simulation study.