Amartya Sanyal

This article examines how preprocessing steps can impact data privacy guarantees.

Exploring how noisy data affects model performance on unseen data.

A method for estimating means despite the impact of outliers.

New method BaDLoss enhances protection against data poisoning in machine learning.

Learn how machine unlearning aids data privacy and model accuracy.

A new method to detect and fix data poisoning in machine learning models.