2nd Workshop on Formal Verification of Machine Learning (WFVML 2023)

About This Workshop

Our workshop accepted 23 high quality papers. See list of accepted papers.

As machine learning-based systems are being deployed in safety-critical applications such as autonomous driving, medical imaging, or cyber-security systems, characterizing their behavior not only in the average but also worst case becomes essential. However, most existing research treats machine learning models such as deep neural networks as black boxes and uses simple empirical metrics such as their mean accuracy to quantify their performance. However, accuracy alone is not sufficient to assure that models conform to even basic safety or robustness specifications. To fill this gap, formal verification algorithms for machine learning aim to formally prove or disprove desired properties of machine learning models, including safety, fault tolerance, fairness, robustness, and correctness. 

The aims of this workshop are:

• Bring together researchers interested in the emerging field of machine learning verification from a broad range of backgrounds (e.g., computer-aided verification, programming languages, robotics and control, and optimization) and with different perspectives on this problem; 

• Raise awareness for the importance of formal verification in the machine learning community, discuss the open challenges defining the field over the next years, and stimulate research tackling these hard open problems;

• Chart out important and promising future directions toward novel verification algorithms with better scalability and applicability.

Our workshop features a diverse panel of invited speakers spanning research backgrounds from robotics and programming languages to optimization and cyber security. Please check out our tentative workshop schedule.