Dense_Associative_Memory

An example of Dense Associative Memory training with a backpropagation algorithm on MNIST. Based on the paper Dense Associative Memory for Pattern Recognition by Dmitry Krotov and John Hopfield. If you want to learn more about Dense Associative Memories, check out a NIPS 2016 talk or a research seminar.

Getting started

install jupyter notebook and numpy, scipy, matplotlib.

> jupyter notebook

run Dense_Associative_Memory_training.ipynb and observe the weights together with the errors on the training and the test sets.

Author and License

(c) 2016 Dmitry Krotov -- Apache 2.0 License