tyrosine
| Crates.io | tyrosine |
| lib.rs | tyrosine |
| version | 0.9.0 |
| created_at | 2023-12-23 22:24:56.19673+00 |
| updated_at | 2025-07-24 00:08:36.781516+00 |
| description | Rust implementation of the NEAT neuroevolution algorithm. |
| homepage | https://github.com/genomicals/tyrosine |
| repository | https://github.com/genomicals/tyrosine |
| max_upload_size | |
| id | 1079306 |
| size | 38,982 |
Opossalite (opossalite)
documentation
README
Tyrosine
Mascot by @Nauxe
Overview
NEAT machine learning library for Rust.
NeuroEvolution of Augmenting Topologies (NEAT) is a genetic algorithm which evolves a population of neural networks. The population is split into several genetically-distinct species. NEAT aims to find an optimal set of weights in addition to an optimal neural network structure.
Link to original paper here.
Implementation Specifics
There are several instances where specific features are up to the implementer's choice. Here are some choices we made:
- A bias node is included in the inputs
- Cyclic connections will be disallowed
- Species won't be removed for becoming stale, only when allocated reproductive slots reach 0
Trivia
Tyrosine is an amino acid. The word comes from the Greek "tyros", which means cheese.
License
This work is dual-licensed under GPL Version 2 and GPL Version 3. You may choose between these licenses on a case-by-case basis.