Papers
10
Total Citations
270
H-Index
8
About
Riad Akrour is a leading researcher at the intersection of reinforcement learning (RL) and robotics, with a particular focus on making robot learning more sample-efficient, interactive, and human-centric. His foundational work on **preference-based reinforcement learning** has been highly influential—his 2012 paper “APRIL: Active Preference Learning-Based Reinforcement Learning” (94 citations) pioneered a paradigm where robots learn from human preferences rather than hand-crafted reward functions, dramatically reducing the burden on experts. This line of work, including “Preference-Based Policy Learning” (83 citations), established Akrour as a key figure in interactive robot education. Akrour has also made significant contributions to **hierarchical reinforcement learning** and **state abstraction**, developing methods that enable robots to learn complex, high-dimensional manipulation tasks. His 2018 paper on regularizing RL with state abstraction (21 citations) generalized abstraction to continuous action spaces, while his work on hierarchical tactile-based control (2020, 17 citations) advanced dexterous in-hand manipulation using tactile feedback. Across his career, Akrour’s research has consistently addressed the core challenge of bridging human intuition and machine learning, creating algorithms that are both theoretically sound and practically deployable on physical robots.
Research Focus
Key Achievements
Top Papers
- 1APRIL: Active Preference Learning-Based Reinforcement Learning94 citations · 2012
- 2Preference-Based Policy Learning83 citations · 2011
- 3Regularizing Reinforcement Learning with State Abstraction21 citations · 2018
- 4
- 5
- 6Layered direct policy search for learning hierarchical skills11 citations · 2017
- 7APRIL: Active Preference-learning based Reinforcement Learning10 citations · 2012
- 8Interactive Robot Education9 citations · 2013
- 9Empowered skills4 citations · 2017
- 10Learning Replanning Policies With Direct Policy Search2 citations · 2019