Skip to main content
Fig. 3 | BMC Medical Research Methodology

Fig. 3

From: Utilization of five data mining algorithms combined with simplified preprocessing to establish reference intervals of thyroid-related hormones for non-elderly adults

Fig. 3

The diagram of algorithms for RIs establishment using the Test data set. Hoffmann and Bhattacharya distinguish the distribution of the healthy individuals from the mixed distribution with the linear region modeling visually. RIs for healthy subgroup are obtained by extending the linear region in these two methods. For Expectation-Maximum algorithm, the diagrams described the distribution of the mixed values after the Box-cox transformation. The green curve denotes the distribution of the healthy population while the blue and red curve represent the distribution of the pathological population. For kosmic, the distribution curve that stands for the estimated distribution is used for establishing RIs. For refineR, the distribution of the healthy population is denoted as green curve, with the green vertical dotted line indicating the limits of the RIs and the shaded vertical areas indicating the 90% confidence interval

Back to article page