In caranha/EvoGCP: Evolutionary Computation Tools for the Graph Coloring Problem

The 3GCP package implements a number of 3-Graph Coloring Problem solvers, focusing on swarm-based meta-heuristics. The algorithms are implemented on a component-oriented format, so that the user can freely mix and match components from different algorithms.

The package provides support for evaluating one or more algorithms, for automated fine-tuning algorithm parameters, and for auto-discovery of algorithm component composition.

Package Components

Definition of a 3-GCP problem. Package provides generation and evaluation. (Generation, Random, and Paper)

Definition of a swarm-based meta-heuristic solver.

Definition of the fine-tuning component.

Definition of the evaluation/comparison component.

Definition of the auto-discovery component.

Algorithm Outline and Component Classes

Individual

Initial Generation

Variation

Evaluation

Selection

Component Details

For each class above, list the components implemented.

Base Algorithm-sets

• Aoki's Hamming-PSO
• Toda's Cuckoo Search
• Togashi's ABC
• Full random Algorithm

caranha/EvoGCP documentation built on May 3, 2021, 3:40 p.m.