Lyle Ungar

About

Lyle Ungar is a pioneering researcher in machine learning and robotics, best known for his foundational work in active learning and reinforcement learning for autonomous systems. His research focuses on developing algorithms that enable robots to efficiently learn and adapt in real-world environments, particularly through active exploration and decision-making under uncertainty. Ungar’s major contributions include the introduction of active learning techniques for vision-based robot grasping, as demonstrated in his highly cited 1996 paper (53 citations), which showed how robots could autonomously select informative training examples to improve grasping performance. He also advanced policy gradient reinforcement learning for robot controllers (20 citations), enabling robust performance in complex, dynamic settings. Additionally, his work on multi-armed bandit allocation indices (16 citations) provided a theoretical framework for active exploration in real-valued spaces. With over 100 total citations across his most influential works, Ungar’s research has had a lasting impact on robotics and machine learning, bridging theory and practice to create more intelligent, autonomous systems. His achievements continue to inspire students and researchers in AI and robotics.

Research Focus

Key Achievements

Top Papers

1. 1
   Active learning for vision-based robot grasping

   53 citations · 1996
2. 2
   Using policy gradient reinforcement learning on autonomous robot controllers

   20 citations · 2004
3. 3
   Active Exploration and Learning in Real-Valued Spaces using Multi-Armed Bandit Allocation Indices

   16 citations · 1995
4. 4
   Active Learning for Vision-Based Robot Grasping

   6 citations · 1996

Key Collaborators

Contact & Links