Metadata-Version: 2.2
Name: decisionbandit
Version: 0.1.0
Summary: A collection of exploration–exploitation strategies for reinforcement learning, including ε-greedy and related policies
Author: Adit Punamiya
Author-email: adit@gmail.com
License: MIT
Requires-Python: >=3.8
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: numpy
Dynamic: author
Dynamic: author-email
Dynamic: description
Dynamic: description-content-type
Dynamic: license
Dynamic: requires-dist
Dynamic: requires-python
Dynamic: summary

# epsilon-policies

A collection of explorationâ€“exploitation strategies for reinforcement learning, including Îµ-greedy and related policies.

## Features
- **Exploration** â€“ Select random actions to discover new possibilities.
- **Exploitation** â€“ Choose the best-known action based on current estimates.
- **Fixed Explorationâ€“Thenâ€“Exploitation** â€“ Explore for a fixed period, then fully exploit.
- **Îµ-Greedy** â€“ Balance exploration and exploitation with a probability parameter.
- **Îµ-Greedy with UCB** â€“ Enhance Îµ-greedy with Upper Confidence Bound for better action selection.

## Installation


```bash
pip install decisionbandit


import decisionbandit as dcb

# Example: Îµ-greedy
action = dcb.epsilon_greedy(q_values=[1.0, 0.5, 0.2], epsilon=0.1)
print("Selected action:", action)


MIT License
