eXplainable and reliable against adversarial machine learning

Abstract— Machine learning models are increasingly being integrated into critical decision-making processes across various domains. However, these models are susceptible to adversarial attacks, where malicious actors deliberately manipulate input data to deceive the models and induce incorrect predictions. In this paper, we present an overview of state-of-the-art techniques that aim to enhance the explainability and reliability of machine learning models in the face of adversarial attacks. We begin by discussing the fundamental concepts and motivations behind adversarial machine learning, emphasizing the need for models that can provide explanations for their predictions while maintaining robustness. Keywords— Explainability,Reliability,Adversarial attacks Robustness.