self-play (adversarial training)

1 episode mention this concept

Related concepts — line thickness indicates connection strength. Click any node to explore.

Related concepts

Concepts that appear alongside self-play (adversarial training) across episodes.

• hierarchical reinforcement learning
• sparse reward problem
• sim-to-real transfer (with randomization)
• sample efficiency (data efficiency)
• computational intractability (program search)
• meta-learning paradigms (learning to learn)
• policy gradient algorithms
• representation learning (for goals)
• hindsight experience replay (her)
• backpropagation (circuit search)
• generalization theory (shortest program)
• deep learning principles
• model-free rl (randomness, expectation)
• reinforcement learning framework
• q-learning algorithms