Daniel J. Mankowitz

About

Daniel J. Mankowitz is a leading researcher in reinforcement learning (RL), with a primary focus on developing robust algorithms for continuous control systems. His most impactful work addresses a critical challenge in real-world RL deployment: model misspecification, where the simulated environment differs from reality. In his highly cited 2019 paper (38 citations), Mankowitz introduced a novel framework that incorporates robustness against perturbations in transition dynamics into state-of-the-art continuous control algorithms. This contribution is essential for deploying RL in safety-critical domains like robotics and autonomous systems, where small environmental changes can lead to catastrophic failures. Beyond robustness, his earlier work on BRISK-based visual feature extraction (2014) demonstrates his versatility in tackling resource-constrained robotic applications, optimizing computer vision for platforms with limited computational power. Mankowitz's research bridges the gap between theoretical RL advances and practical deployment, making his work foundational for students and engineers seeking to build reliable, real-world learning systems. His focus on robustness continues to influence modern RL research, particularly in areas requiring safe and adaptive decision-making under uncertainty.

Research Focus

Key Achievements

Top Papers

1. 1
   Robust Reinforcement Learning for Continuous Control with Model Misspecification

   38 citations · 2019
2. 2
   Robust Reinforcement Learning for Continuous Control with Model Misspecification

   6 citations · 2020
3. 3
   BRISK-Based Visual Feature Extraction for Resource Constrained Robots

   5 citations · 2014

Key Collaborators

Contact & Links