Safe and Robust Binary Classification and Fault Detection Using Reinforcement Learning

In this paper, we propose a learning-based method utilizing the Soft Actor-Critic (SAC) algorithm to train a binary Support Vector Machine (SVM) classifier. This classifier is designed to identify valid input spaces in high-dimensional, highly constrained systems while minimizing the total runtime o...

Full description

Saved in:
Bibliographic Details
Main Authors: Josh Netter, Kyriakos G. Vamvoudakis, Timothy F. Walsh, Jaideep Ray
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Open Journal of Control Systems
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11010134/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849432905861824512
author Josh Netter
Kyriakos G. Vamvoudakis
Timothy F. Walsh
Jaideep Ray
author_facet Josh Netter
Kyriakos G. Vamvoudakis
Timothy F. Walsh
Jaideep Ray
author_sort Josh Netter
collection DOAJ
description In this paper, we propose a learning-based method utilizing the Soft Actor-Critic (SAC) algorithm to train a binary Support Vector Machine (SVM) classifier. This classifier is designed to identify valid input spaces in high-dimensional, highly constrained systems while minimizing the total runtime of offline simulations. The simulations adapt their runtime based on the likelihood that a given training input will be informative to the classifier. Furthermore, we introduce a method for using the trained SAC model to predict whether a desired system input is likely to violate constraints, along with a technique to adjust the input as necessary. Additionally, we explore the potential of this model to detect faults or adversarial attacks within the system. The effectiveness of our approach is demonstrated through various simulations of challenging classification problems and a constrained quadrotor model.
format Article
id doaj-art-e72589d3a6eb4f20b97f92c96d460e38
institution Kabale University
issn 2694-085X
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Open Journal of Control Systems
spelling doaj-art-e72589d3a6eb4f20b97f92c96d460e382025-08-20T03:27:14ZengIEEEIEEE Open Journal of Control Systems2694-085X2025-01-01417218610.1109/OJCSYS.2025.357237511010134Safe and Robust Binary Classification and Fault Detection Using Reinforcement LearningJosh Netter0https://orcid.org/0009-0001-0566-9024Kyriakos G. Vamvoudakis1https://orcid.org/0000-0003-1978-4848Timothy F. Walsh2https://orcid.org/0000-0002-9527-4025Jaideep Ray3https://orcid.org/0009-0000-9908-7035Daniel Guggenheim School of Aerospace Engineering, Georgia Institute of Technology, Atlanta, GA, USADaniel Guggenheim School of Aerospace Engineering, Georgia Institute of Technology, Atlanta, GA, USASandia National Laboratories, Albuquerque, NM, USASandia National Laboratories, Albuquerque, NM, USAIn this paper, we propose a learning-based method utilizing the Soft Actor-Critic (SAC) algorithm to train a binary Support Vector Machine (SVM) classifier. This classifier is designed to identify valid input spaces in high-dimensional, highly constrained systems while minimizing the total runtime of offline simulations. The simulations adapt their runtime based on the likelihood that a given training input will be informative to the classifier. Furthermore, we introduce a method for using the trained SAC model to predict whether a desired system input is likely to violate constraints, along with a technique to adjust the input as necessary. Additionally, we explore the potential of this model to detect faults or adversarial attacks within the system. The effectiveness of our approach is demonstrated through various simulations of challenging classification problems and a constrained quadrotor model.https://ieeexplore.ieee.org/document/11010134/Machine learningreinforcement learningrobust control
spellingShingle Josh Netter
Kyriakos G. Vamvoudakis
Timothy F. Walsh
Jaideep Ray
Safe and Robust Binary Classification and Fault Detection Using Reinforcement Learning
IEEE Open Journal of Control Systems
Machine learning
reinforcement learning
robust control
title Safe and Robust Binary Classification and Fault Detection Using Reinforcement Learning
title_full Safe and Robust Binary Classification and Fault Detection Using Reinforcement Learning
title_fullStr Safe and Robust Binary Classification and Fault Detection Using Reinforcement Learning
title_full_unstemmed Safe and Robust Binary Classification and Fault Detection Using Reinforcement Learning
title_short Safe and Robust Binary Classification and Fault Detection Using Reinforcement Learning
title_sort safe and robust binary classification and fault detection using reinforcement learning
topic Machine learning
reinforcement learning
robust control
url https://ieeexplore.ieee.org/document/11010134/
work_keys_str_mv AT joshnetter safeandrobustbinaryclassificationandfaultdetectionusingreinforcementlearning
AT kyriakosgvamvoudakis safeandrobustbinaryclassificationandfaultdetectionusingreinforcementlearning
AT timothyfwalsh safeandrobustbinaryclassificationandfaultdetectionusingreinforcementlearning
AT jaideepray safeandrobustbinaryclassificationandfaultdetectionusingreinforcementlearning