Anirudh Suresh

Papers

Total Citations

H-Index

About

Anirudh Suresh is a researcher whose work lies at the intersection of Bayesian machine learning, uncertainty quantification, and interpretable AI. His most notable contribution is the development of Output-Constrained Bayesian Neural Networks (OC-BNNs), a framework that addresses a fundamental limitation of standard BNNs: the difficulty of encoding prior knowledge directly in function space. By formulating priors that constrain model outputs in specific regions of the input space, Suresh enables practitioners to incorporate domain expertise—such as physical laws or safety constraints—directly into neural network training. This innovation, detailed in his 2019 paper (10 citations), has significant implications for high-stakes applications like autonomous systems and healthcare, where model outputs must adhere to known boundaries. Beyond this, his research explores robust and trustworthy machine learning, aiming to bridge the gap between probabilistic modeling and real-world deployment. Suresh’s work is particularly valuable for students and researchers seeking to build models that are not only accurate but also reliable and aligned with prior knowledge, making him a key voice in the push toward safer, more interpretable AI systems.