Development of a Software System for a Neural Net-work Study of the Phenomenon of Depression

Author of the work: Kokovin Alexey Nikolaevich, SE-16-2, HSE-Perm. Title: Implementation of a software system for a neural network study of the phenomenon of depression. Number of Pages: 40. Number of pictures: 5. Number of tables: 2. In Chapter 1 "A Survey on Research in the Diagnosis of Depression and the Socio-Biological Factors Causing Its Development" research papers on socio-biological factors of the depression and modern methods of diagnosis of the depression are compiled in order to confirm usability of data and methods of previous study and existing software solutions are analysed to pinpoint software Chapter 2 “Designing affordable systems and creating technical specifications” describes the requirements put forward by the system, and also designes the functionality of the diagnostic system for depression and their implementation. Chapter 3 “Implementing and testing an access system” describes implementation of programm system: a set of components, implementation details of training component for a neural network and a web server, test coverage of the code. Chapter 4 “Evaluation and study of neural network models” describes the characteristics of the resulting neural network models, which are used to draw conclusions at the level of factors affecting the development of depression. The work contains 7 applications. This work may interest data scientists interested in the application of machine learning methods and, in particular, a neural network for diagnosing mental disorders, in particular depression, or to professional psychiatrists interested in studying the influence of sociobiological factors on the occurrence of depression.