Determination of radon prone areas by optimized binary classification

The overall results are different maps of radon prone areas, derived from the geogenic radon map, and depending (1) on the criterion which defines what a radon prone area is; and (2) on the choice of score whose maximization defines the optimal classifier. Such map is not the result of a transfer model (geogenic to indoor radon), but of the optimization of a classification rule. The method is computationally simple but has its caveats and statistical traps, some of which are also addressed.