takuseno/minerva

An out-of-the-box GUI tool for offline deep reinforcement learning

37
/ 100
Emerging

This tool helps anyone, even non-programmers, use advanced deep reinforcement learning to solve problems. You provide a dataset of observations and actions, and it trains a powerful decision-making policy. The output is a trained policy that can be exported for use in other systems. It's designed for data scientists, researchers, or practitioners looking to leverage AI for automated decision-making.

102 stars. No commits in the last 6 months.

Use this if you have existing interaction data and want to train an AI to make optimal decisions based on that data, without needing to write code.

Not ideal if you need a reinforcement learning system that interacts directly with an environment in real-time or if you require extensive custom algorithm development.

AI-driven decision making Offline reinforcement learning Policy optimization Behavior cloning Automated strategy
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 9 / 25
Maturity 16 / 25
Community 12 / 25

How are scores calculated?

Stars

102

Forks

10

Language

JavaScript

License

MIT

Last pushed

May 29, 2021

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/takuseno/minerva"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.