|
--- |
|
title: Submission Template |
|
emoji: 🔥 |
|
colorFrom: yellow |
|
colorTo: green |
|
sdk: docker |
|
pinned: false |
|
--- |
|
|
|
|
|
# Random Forest Model for Climate Disinformation Classification |
|
|
|
## Model Description |
|
|
|
This is a random forest model for the Frugal AI Challenge 2024, specifically for the audio classification task of identifying illegal deforestation. |
|
|
|
### Intended Use |
|
|
|
- **Primary intended uses**: Illegal deforestation classification model |
|
- **Primary intended users**: Researchers and developers participating in the Frugal AI Challenge |
|
- **Out-of-scope use cases**: Not intended for production use or real-world classification tasks |
|
|
|
## Training Data |
|
|
|
The model uses the rfcx/frugalai dataset: |
|
- Size: ~50000 examples |
|
- Split: 80% train, 20% test |
|
- 2 categories of audio category |
|
|
|
### Labels |
|
0. chainsaw (positively identifying a chainsaw) |
|
1. environment (not containing a chainsaw). |
|
|
|
## Performance |
|
|
|
### Metrics |
|
- **Accuracy**: ~89.3% |
|
- **Environmental Impact**: |
|
- Emissions tracked in gCO2eq |
|
- Energy consumption tracked in Wh |
|
|
|
### Model Architecture |
|
The model implements a random forest model used on pre-processed data. The pre-processing consists in a resampling, a Fourier decomposition and a standard scaler. |
|
|
|
## Environmental Impact |
|
|
|
Environmental impact is tracked using CodeCarbon, measuring: |
|
- Carbon emissions during inference |
|
- Energy consumption during inference |
|
|
|
This tracking helps establish a baseline for the environmental impact of model deployment and inference. |
|
|
|
## Limitations |
|
- Takes some time to do the pre-processing. |
|
|
|
## Ethical Considerations |
|
|
|
- Environmental impact is tracked to promote awareness of AI's carbon footprint |
|
``` |
|
|
