https://github.com/srshtiputcha/sgmcmc_preferential_subsampling

Code for Preferential Subsampling for Stochastic Gradient Langevin Dynamics
https://github.com/srshtiputcha/sgmcmc_preferential_subsampling

Last synced: 27 days ago
JSON representation

Code for Preferential Subsampling for Stochastic Gradient Langevin Dynamics

• Host: GitHub
• URL: https://github.com/srshtiputcha/sgmcmc_preferential_subsampling
• Owner: srshtiputcha
• License: mit
• Created: 2022-10-10T14:12:39.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2022-10-31T08:49:49.000Z (over 1 year ago)
• Last Synced: 2024-02-17T23:32:09.774Z (4 months ago)
• Language: Jupyter Notebook
• Homepage: https://arxiv.org/abs/2210.16189
• Size: 2.11 MB
• Stars: 3
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Lists

• awesome-stars - srshtiputcha/sgmcmc_preferential_subsampling - Code for Preferential Subsampling for Stochastic Gradient Langevin Dynamics (Jupyter Notebook)

README

        
# sgmcmc_preferential_subsampling

### Code for Preferential Subsampling for Stochastic Gradient Langevin Dynamics

For full details, please refer to the [arXiv preprint](https://arxiv.org/abs/2210.16189) of this work. 

All code is written in `Python 3.7x` and the key list of module requirements are provided in `requirements.txt`.

Run `./setup.sh` to set up the `data/` folder. This will generate all synthetic data and download the real data used in this paper. 

The source code for both models and samplers can be found in the `src/` folder. 

All experiments can be found in the `experiments/` folder. Jupyter Notebooks for each experiment should be opened within their containing folder. Each experiment folder contains a `Description.txt` file which details the purpose of each notebook.