Using the isolation forest algorithm to solve the problem of detecting anomalies in the operation of microprocessor plastic cards

Abstract

Anomaly detection is an important issue in data science. The purpose of detection is to determine the system and stochastic violations of the signal structure, the regularity of repetition of characteristic sites, as well as the localization of artifacts of natural or artificial origin. These can be failures in the data collection process, sensor failure, or external unexpected events, such as an intentional violation of the microcontroller’s security. In addition, with the help of the proposed methods of anomaly detection, it is possible not only to analyze the security of devices, but also to determine the efficiency of the use of computing resources and check the operation of hardware modules. In General, the introduction of such methods is advisable at all stages of the life cycle of microprocessors and microcontrollers.