Skip to content
This repository has been archived by the owner. It is now read-only.
An OpenAI training gym environment for MNIST handwritten digit classification with a Deep-Q Agent and CNN Policy. Digits 0-9 are displayed on a upsampled 128x128px canvas. A correct discrete value action for a matching observation receives a reward.
Jupyter Notebook
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.gitignore
LICENSE
README.md
dqn_cnn_mnist.ipynb
dqn_cnn_mnist.zip

README.md

dqn_cnn_mnist_gym

An OpenAI training gym environment for MNIST handwritten digit classification with a Deep-Q Agent and CNN Policy. Digits 0-9 are displayed on a upsampled 128x128px canvas. A correct discrete value action for a matching observation receives a reward.

Screenshot

This model achieves a 98% accuracy on the MNIST dataset.

You can’t perform that action at this time.