https://github.com/vida-nyu/gdpfinder
Python code to train and evaluate machine learning models for the estimation of neighborhood-level census statistics.
https://github.com/vida-nyu/gdpfinder
Last synced: 3 months ago
JSON representation
Python code to train and evaluate machine learning models for the estimation of neighborhood-level census statistics.
- Host: GitHub
- URL: https://github.com/vida-nyu/gdpfinder
- Owner: VIDA-NYU
- Created: 2023-02-24T18:26:59.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2024-02-26T21:46:33.000Z (over 1 year ago)
- Last Synced: 2025-01-24T15:36:42.729Z (5 months ago)
- Language: Jupyter Notebook
- Homepage:
- Size: 223 MB
- Stars: 0
- Watchers: 9
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Granularity at Scale
Paper: [Granularity at Scale: Estimating Neighborhood Socioeconomic Indicators from High-Resolution Orthographic Imagery and Hybrid Learning](https://doi.org/10.1109/JSTARS.2024.3368018)
## For the supervised approach:
```notebooks/supervised_approach.ipynb``` contains data (census and satellite imagery) preparation and analysis. Cells to test saved models and interpret results are also there.
### Example usage to train a model:
Within ```scripts``` directory:
```
$ nohup python -u supervised_training.py --metric 'density' --imagetype 'resize' --newwidth 1234 --newheight 1234 &
```
See ```supervised_training.py``` for more details on the arguments and training process.
In ```supervised_approach.ipynb``` and ```supervised_training.py```, the dataset is generated from ```scripts/create_dataset.py``` and the model from ```scripts/supervised_models.py```