Nikolai Matni

A novel approach improves representation learning across various tasks using de-biasing and feature-whitening techniques.

Researchers are improving how robots learn through human demonstrations, enhancing safety and adaptability.

Learn the process of creating models for dynamic systems using measured data.

Strategies for effectively learning from data that depends on previous observations.

A method to efficiently model and control nonlinear systems using active learning techniques.

New method enhances robots' ability to work safely alongside humans in real-time.

This study explores how representation learning helps robots adapt in changing environments.

A new approach to improve coordination in layered control systems using reinforcement learning.

Examining how model parameters impact performance on long-range tasks.