Optimization of features in the problem of recognizing a person’s gender by his handwriting

Myasoutov Ramil Khamzyaevich

In the task of recognizing the gender of a person by his handwriting the characteristics contained in any studied sample play an extremely important role. At the moment 3193 features have been collected that describe a person's handwriting. It is rather difficult to work with such a large number of features since their search and analysis takes a lot of time. The goal of this work is to optimize the number of features so as to leave the smallest number of features at which the accuracy of sex determination remains acceptable.

Keywords:feature optimization, dimensionality reduction, classification problem.