Intermediate · Reinforcement Learning
Epsilon greedy policy
Visual diagram · (in preparation) · Math · (in preparation) · Worked example · 3 difficulty levels.
TL;DR. A strategy where an agent randomly explores the environment most of the time, but sometimes exploits known good actions.
Technical Definition
A strategy where an agent randomly explores the environment most of the time, but sometimes exploits known good actions.
How it works
The epsilon-greedy policy is a common exploration strategy in reinforcement learning. With probability epsilon, the agent chooses a random action to explore new possibilities, and with probability 1-epsilon, it chooses the action it currently believes to be the best (greedy action). Epsilon is often decreased over time to favor exploitation as the agent learns.
Related Concepts
- Reinforcement Learning — A paradigm where an agent learns to make decisions by receiving rewards or penalties from its environment through trial and error.
- Policy — In reinforcement learning, the strategy an agent follows to choose actions given a state.