Statistical classifier for determining the localization area of oncopathology using a blood test

A mathematical approach is described that makes it possible to determine the location of the tumor based on the analysis of peripheral blood of a cancer patient. A statistical classifier developed by the authors is used for the preliminary diagnosis, based on polynomial regression and having probabilistic estimates. The classifier was trained on peripheral blood databases of male cancer patients in seven different body systems. A method for studying the structure of the training set is proposed and implemented.