https://github.com/tmcclintock/latinhypercubesampling

An example of using a Latin hypercube to search a hyperparamter space.
https://github.com/tmcclintock/latinhypercubesampling

Last synced: 14 days ago
JSON representation

An example of using a Latin hypercube to search a hyperparamter space.

• Host: GitHub
• URL: https://github.com/tmcclintock/latinhypercubesampling
• Owner: tmcclintock
• License: mit
• Created: 2020-08-17T19:47:27.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2020-08-21T01:21:13.000Z (over 4 years ago)
• Last Synced: 2024-12-14T17:09:22.790Z (about 1 month ago)
• Language: Jupyter Notebook
• Size: 35.2 KB
• Stars: 0
• Watchers: 2
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# LatinHyperCubeSampling

This repository is for a yet to be written article about using a Latin

hypercube for randomly sampling hyperparameters in a deep learning setting. This

is advantageous over a grid for scaling reasons, and is better than purely

random sampling since that is not space filling.

## Requirements

You can install the requirements for performing the experiments here with the

`requirements.txt` file as follows (I recommend you do so in an [`anaconda`

environment](https://docs.conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html):

```bash

pip install -r requirements.txt

```