Understanding Memory-Equivalent Capacity in Unsupervised Learning

Implementation of Understanding Memory-Equivalent Capacity in Unsupervised Learning

Getting Started

Run

  python theoretical_vs_empirical_LM_bounds_main.py

to run capacity experiments.

Results

The following results were obtained for LM and MK capacities.

Note that empirical results are affected by the low sample size for label configurations.

Authors

• Jin Park - @jpark96

License

This project is licensed under the MIT License.

Acknowledgments