A tutorial on Yandex studcamp in Innopolis 2024

The tutorial aims to show how to apply adversarial attacks to fault diagnosis systems and how to protect them using the adversarial training technique. As a result, students will see the vulnerability of fault diagnosis systems to adversarial attacks and learn how to protect such systems.

The target audience is undergraduate computer science students who want to explore the robustness of fault diagnosis systems. I assume that the audience has a basic knowledge of machine learning, probability theory, statistics and Python programming.

Open jupyter notebook

Topics: Tennessee Eastman Process, Fault Diagnosis Systems, Multi-Layer-Perceptron, Adversarial Attacks, Fast Gradient Sign Method, One-step Target Class Method, Adversarial Training.

Vitaliy Pozdnyakov junior research scientist, AIRI

Vitaliy Pozdnyakov has 7 years of experience in industrial companies as a developer of enterprise resource planning systems and 3 years in scientific research of industrial artificial intelligence methods. Research interests: generative models for time series applications, graph neural networks for operational research, self-supervised learning for fault diagnosis, generative models for risk management.