Soban Nasir Lone

Soban Nasir Lone

GNNs for Traffic Prediction

Publications

You can also find my articles on my Google Scholar profile.

Recursive Bayesian neural networks

Published in arXiv preprint, 2024

This study proposes a recursive Bayesian neural network (rBNN) framework for uncertainty-aware constitutive modelling in geotechnical engineering, incorporating a sliding window approach to capture temporal dependencies. Validated on numerical and experimental triaxial datasets, the rBNN provides robust confidence intervals, highlighting trade-offs between deterministic and probabilistic models.

Recommended citation: Noor, T., Nasir Lone, S., Ramana, G.V. and Nayek, R., 2025. A recursive Bayesian neural network for constitutive modeling of sands under monotonic loading. arXiv e-prints, pp.arXiv-2501.
Download Paper

Alpha-VI DeepONet

Published in arXiv preprint, 2024

We introduce a novel deep operator network (DeepONet) framework that incorporates generalised variational inference (GVI) using Rényi’s α-divergence to learn complex operators while quantifying uncertainty. We apply this approach to a range of mechanics problems, including gravity pendulum, advection-diffusion, and diffusion-reaction systems.

Recommended citation: S. N. Lone, S. De, R. Nayek, Alpha-VI DeepONet: A prior-robust variational Bayesian approach for enhancing DeepONets with uncertainty quantification, arXiv preprint arXiv:2408.00681 (2024)
Download Paper