Розробка методу та програмного зaсобу ідентифікації користувача системи безпеки за клавіатурним почерком на основі нейромереж

Abstract

With the ever-increasing systemic threats and security requirements to prevent these threats, there is a need for reliable security systems. Password-based identification systems are very vulnerable to attacks. Keyboard user identification is considered one of the alternative, modern, reliable solutions due to its low cost and reliability, non-intrusiveness and user-friendliness. The master's degree work is devoted to the development of a method and a software tool for user identification of a security system by keyboard-based handwriting based on neural networks. In the course of work the analysis of existing neural networks was carried out. The neural network architecture was designed. A software tool for identifying the security user by keyboard is developed.