|
|
--- |
|
|
tags: |
|
|
- health |
|
|
- stroke-risk |
|
|
- machine-learning |
|
|
- gradio |
|
|
library_name: gradio |
|
|
license: mit |
|
|
widget: |
|
|
- type: gradio |
|
|
src: app.py |
|
|
--- |
|
|
|
|
|
|
|
|
|
|
|
# Stroke Risk Prediction Model π |
|
|
|
|
|
This model predicts the stroke risk percentage based on user symptoms using a trained linear regression model. |
|
|
|
|
|
## π Features: |
|
|
- β
Takes 16 symptoms as input (Checkbox selection) |
|
|
- β
Returns a stroke risk percentage |
|
|
- β
Deployed using Gradio on Hugging Face Spaces |
|
|
|
|
|
## π§ How It Works: |
|
|
1. User selects relevant symptoms. |
|
|
2. The input is normalized based on precomputed dataset statistics. |
|
|
3. The trained model (`theta_final.npy`) predicts the stroke risk. |
|
|
|
|
|
## π Try it Live: |
|
|
[](https://huggingface.co/attiquers) |
|
|
|
|
|
## π Files: |
|
|
- `app.py`: Gradio interface and model inference. |
|
|
- `theta_final.npy`: Trained model parameters. |
|
|
- `requirements.txt`: Dependencies. |
|
|
|
|
|
## π Installation (Local Testing): |
|
|
```bash |
|
|
pip install gradio numpy |
|
|
python app.py |
|
|
|