The efficiency of logistic function and prediction-area plot in prospectivity analysis of mineral deposits.

Abstract

In this work, we present logistic-based mineral prospectivity mapping (MPM) methods concerning with assigning weights of exploration indicators, without contribution of training sites as in supervised MPM and without using user-judged weights as in unsupervised MPM, to modulate the problems of stochastic and systemic errors. In addition, we discuss the ability of prediction-area plot as a tool to assess and compare evidential layers and prospectivity models.