Saurabh Sihag

Assistant Professor at University at Albany.

I am an Assistant Professor in the Electrical and Computer Engineering department at State University of New York at Albany. Previously, I was a postdoctoral researcher working with Dr. Alejandro Ribeiro at the University of Pennsylvania. Broadly, I study statistical inference and machine learning approaches over graph models, both from theoretic and algorithmic perspectives with applications in network neuroscience. I had received the PhD degree in Electrical Engineering at Rensselaer Polytechnic Institute in Dec., 2020, where I was advised by Dr. Ali Tajer. My PhD thesis on Statistical Learning and Inference over Networks can be accessed here.

My research is focused on developing novel principles of machine learning and statistical inference using concepts from signal processing, learning theory, information theory, and graph theory. I have worked on a variety of research problems that include statistical learning of graph models (AISTATS, NeurIPS), state estimation in signal processing (T-IT, JSTSP, SPL), graph signal processing analyses of multimodal brain imaging data (TSIPN) and association of neuroimaging features with blood biomarkers (NatCommsMed), and adaptive graph-constrained group testing (TSP). Currently, I am investigating statistical inference using graph neural networks (VNN), with brain age prediction as a recent application (VNN_Brain_Age). A complete list of my research publications is available here.

news

May, 2024 Paper on 'Neural Tangent Kernels Motivate Cross-Covariance Graphs in Neural Networks' accepted at ICML, 2024.
March, 2024 Paper on 'Transferability of coVariance Neural Networks' accepted at IEEE Journal of Selected Topics in Signal Processing (JSTSP) Special Series on AI in Signal & Data Science.
September, 2023 Paper on 'Explainable Brain Age Prediction using coVariance Neural Networks' accepted at NeurIPS, 2023.
June, 2023 Paper on 'Learning Graph Structure from Convolutional Mixtures' accepted at Transactions on Machine Learning Research.
June, 2023 Paper on 'Predicting Brain Age using Transferable coVariance Neural Networks' was presented at ICASSP, 2023.
Sep. 14, 2022 Paper on 'coVariance Neural Networks' was accepted at NeurIPS, 2022.
Sep. 1, 2022 Paper on 'Estimating structurally similar graphical models' was accepted at IEEE Transactions on Information Theory.
April, 2022 Our paper on 'Summary Markov Models for Event Sequences' accepted for long presentation at International Joint Conference on Artificial Intelligence, 2022.
April, 2022 Abstract 'Structural Modularity is a Feature of C9orf72 Intrinsic Network Degeneration' accepted for oral presentation at Alzheimer's Association International Conference, 2022.
Dec 10, 2021 Our paper on Adaptive Graph-constrained Group Testing was accepted at IEEE Transactions on Signal Processing.
Dec 7, 2021 Our paper on Functional Brain Signals Constrained by Structural Brain Connectivity Reveal Cohort Specific Features for Serum Neurofilament Light Chain was accepted at Communications Medicine.
Nov 9, 2021 Abstract on De-differentiation of functional networks in corticobasal syndrome was presented virtually at Society for Neuroscience, 2021.