معرفی
Matthew Wicker is an Assistant Professor (Lecturer) at Imperial College London and a Research Associate at The Alan Turing Institute. His research focuses on developing formally verified, trustworthy machine learning systems with provable guarantees for safety, robustness, and fairness. He collaborates extensively with academic and industry partners, including Accenture.
Research Interests:
- Formal verification of ML systems (e.g., neural network robustness/fairness)
- Bayesian neural networks and uncertainty quantification
- Adversarial attack/defense strategies
- Causal structure learning with probabilistic methods
- Certifiable guarantees for model explanations
His recent publications (ICLR, ICML, NeurIPS) converge on theoretical and practical frameworks for trustworthy AI, emphasizing adversarial robustness in Bayesian models, global fairness certification, and scalable uncertainty methods. A consistent theme is bridging formal methods with real-world ML deployment challenges.
Awards:
- Best Paper at TPM 2022 for causal structure learning research
He maintains the deepbayes repository, enabling reproducibility of his PhD work on Bayesian neural network robustness.
