Rust-ML

A simple machine learning library written in Rust, for supervised, unsupervised and reinforcement learning.

Contents

• Neuron: A simple neural network library.
• RL: Implementations of RL algorithms and training environments.
• Examples: Solving classic ML problems using rust-ml (--release recommended)

Neuron

CPU based neural network library implemented using ndarray.

It won't outperform Tensorflow but it should still be very fast. Especially when compiled in --release mode!

RL

Reinforcement learning library containing Agents, Learners and Environments.

A Learner teaches an Agent to master an Environment. All agents, learners and environments are designed to be easily swappable. For example a QAgent can interact with a Jump environment and learn using a QLearner, and that same agent can interact with a Bird environment and learn using a NeuroEvolutionLearner.