Saumitra Mishra

ati
Saumitra Mishra

Saumitra's current research includes interpretable machine learning (IML) that involves designing algorithms for analysing the behaviour of machine learning models. He is interested in both aspects of IML: developing inherently interpretable models, and developing methods to explain black-box models. During his PhD, he has applied IML methods to analyse machine learning models trained for applications in audio (speech, music) and computer vision. His other research interests include adversarial machine learning, deep generative models, and fairness in machine learning.