Using artificial neural networks for analysis of gender differences in younger teenagers

Abstract

Using of neural networks for psychodiagnostic data processing is proposed. The main features of the proposed algorithm are its visibility and highest definiteness in process of training and neural networks using. For this purpose, the structure of neural networks is rigidly connected with initial analyzed data, and training is achieved by their use. Thus a simple feedforward backpropagation network is built. The networks training and testing are carried out by the example of younger teenager's psychodiagnostic data (a total of 111 schoolchildren). It is shown that the proposed algorithm allows to identify psychological traits that are important to assess gender differences efficiently.