Joseph Modayil
E736831
Joseph Modayil is a researcher in artificial intelligence and reinforcement learning, known for co-authoring the influential Rainbow DQN algorithm that combines multiple DQN improvements into a single unified agent.
All labels observed (1)
| Label | Occurrences |
|---|---|
| Joseph Modayil canonical | 1 |
Statements (18)
| Predicate | Object |
|---|---|
| instanceOf |
computer scientist
ⓘ
researcher ⓘ |
| algorithmContribution | Rainbow DQN combines several DQN improvements into one unified agent NERFINISHED ⓘ |
| coAuthored | Rainbow: Combining Improvements in Deep Reinforcement Learning NERFINISHED ⓘ |
| contributedTo | combining multiple DQN extensions into a single agent ⓘ |
| fieldOfWork |
artificial intelligence
ⓘ
deep reinforcement learning ⓘ reinforcement learning ⓘ |
| knownFor | co-authoring the Rainbow DQN algorithm ⓘ |
| notableWork | Rainbow DQN NERFINISHED ⓘ |
| researchInterest |
practical improvements to RL algorithms
ⓘ
scalable reinforcement learning systems ⓘ value-based deep reinforcement learning ⓘ |
| studies |
sample-efficient reinforcement learning methods
ⓘ
stability and performance of deep RL agents ⓘ |
| workFocusesOn |
deep reinforcement learning algorithms
ⓘ
improvements to Deep Q-Networks (DQN) ⓘ value-based reinforcement learning ⓘ |
Referenced by (1)
Full triples — surface form annotated when it differs from this entity's canonical label.