Rolling Horizon Evolutionary Algorithm

An implementation of the Rolling Horizon Evolutionary Algorithm

Installation

using pip

pip install RollingHorizonEA

Usage

To use the rolling horizon evolutionary algorithm, you will need your game class to implement the Environment interface.

Examples

Examples of setting up any game environment can be found in the examples directory and run with:

python run.py

m_max example

num_dims = 600
m = 50
num_evals = 50
rollout_length = 10
mutation_probability = 0.1

# Set up the problem domain as m-max game
environment = MMaxGame(num_dims, m)

rhea = RollingHorizonEvolutionaryAlgorithm(rollout_length, environment, mutation_probability, num_evals)

rhea.run()

Cite

If you want to cite this library, please use the following DOI