Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/andrewjong/estrous-ai
Machine learning models to classify cell phases of the estrous cycle in mice. Research on the reproductive system to contribute to prolonged spaceflight. Project in collaboration with NASA ARC (Space Biosciences, Intelligent Systems).
https://github.com/andrewjong/estrous-ai
ai biology machine-learning nasa science space
Last synced: about 11 hours ago
JSON representation
Machine learning models to classify cell phases of the estrous cycle in mice. Research on the reproductive system to contribute to prolonged spaceflight. Project in collaboration with NASA ARC (Space Biosciences, Intelligent Systems).
- Host: GitHub
- URL: https://github.com/andrewjong/estrous-ai
- Owner: andrewjong
- License: gpl-3.0
- Created: 2018-06-29T18:50:40.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2019-07-25T20:46:49.000Z (over 5 years ago)
- Last Synced: 2024-01-30T04:05:33.328Z (12 months ago)
- Topics: ai, biology, machine-learning, nasa, science, space
- Language: Jupyter Notebook
- Homepage:
- Size: 39.4 MB
- Stars: 1
- Watchers: 2
- Forks: 2
- Open Issues: 10
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Intro
Previous mice experiments found spaceflight may harm ovarian health. More analysis is needed to support prolonged spaceflight. We aim to develop a deep learning algorithm and software to efficiently analyze reproductive-cycle data, and release this tool to the scientific community. **This is a work in progress.**
# Install Dependencies (Anaconda)
We use Anaconda to manage our Python environment. The Python dependencies are listed in `environment.yml`.
To create the conda environment, run:
```bash
conda env create -f environment.yml
```
# Train a Model
The code for a ResNet transfer-learning model currently exists in the repository. You can train it with the following:
```bash
python train.py -e {experiment_name} -d data/[dataset_name] -m {model} {model_params} -o {optimizer} {optimizer_params} -n {num_epochs} -b {batch_size}
```
Use the `-h` flag for help.
By default, training outputs to `experiments/[experiment_name]/[dataset_name]/[model]/`.
# Contributing
To contribute, either talk to the team in person or shoot us an email (Andrew's is [email protected]). To propose new changes, create a new branch and submit a PR.