Ron Meir
Ron Meir -Learning, Decision making & Control in complex dynamic environments
Research Summary
My main focus is on learning, decision making and control in complex dynamic environments. I work on lifelong learning, where an agent continually interacts with an environment and learns to effectively transfer acquired knowledge and skills to facilitate learning new tasks. Situations where a learner works in combination with a ‘teacher’ in forming successful learning curricula are of particular current interest. Within the domain of Reinforcement Learning, I am interested in effective approaches to flexible exploration schemes which subserve both short-term and long-term goals. Applications of these ideas to domains (e.g., medical data) where prior knowledge can be used in order to improve performance are also of interest.
Current Research
- Reinforcement learning in natural and artificial systems
- The perception-action cycle
- Lifelong learning
- Prior knowledge in learning
- The interaction between sensory and motor systems in control
- Learning and representation in deep neural networks