Subhankar Ghosh

I am a graduate student in the Computer Science Department at Washington State University, Pullman, Where I am advised by Professor Yan Yan and Professor Jana Doppa.

Currently I am working on the uncertainty quantification domain(during training and post training), specifically in conformal prediction(CP).(Last updated on 2023 Oct)

Personal Email / Office Email / Google Scholar / Twitter / Github

Research

My accepted papers are highlighted.

	Probabilistically Robust Conformal Prediction Subhankar Ghosh, Yuanjie Shi, Taha Belkhouja, Yan Yan, Janardhan Rao Doppa, Brian Jones UAI, 2023 (Poster Presentation) project page / PDF with Supplementary / GitHub This paper studies the problem of probabilistically robust conformal prediction (PRCP) which ensures robustness to most perturbations around clean input examples.
	NCP: Neighborhood Conformal Prediction Subhankar Ghosh, Taha Belkhouja, Yan Yan, Janardhan Rao Doppa AAAI, 2023 (Oral Presentation) project page / arXiv / GitHub / PPTX We propose a novel algorithm and theory to improve the efficiency of prediction sets for conformal prediction method while maintaining the coverage guarantee.
	A-CZSL: Adversarial Training of Variational Auto-encoders for Continual Zero-shot Learning Subhankar Ghosh IJCNN, 2021 (Oral Presentation) IEEE / arXiv / Github code / video / PPTX We propose a continual zero-shot learning model(A-CZSL) that is more suitable in real-case scenarios to address the issue that can learn sequentially and distinguish classes the model has not seen during training.
	DVGR-CZSL:Dynamic vaes with generative replay for continual zero-shot learning Subhankar Ghosh CVPR, 2021 (findings) CVPR workshop / arXiv / Github code This paper proposes a novel continual zero-shot learning (DVGR-CZSL) model that grows in size with each task and uses generative replay to update itself with previously learned classes to avoid forgetting.

Consider using Leonid Keselman's Jekyll fork of this page.