Results

Model metrics

The model is evaluated using the following metrics:

• Mean Absolute Error.

• Root Mean Squared Error.

• R² score.

The metrics are stored in:

reports/model_metrics.csv

Training history

The training history is stored in:

reports/training_history.csv

This file contains information about the training and validation loss during the training process.

Generated visualizations

The project generates several visualizations in reports/figures/.

Impact probability distribution

Training history

Predictions versus observations

Correlation heatmap

Potential impact timeline

Magnitude versus Palermo scale

Velocity versus Palermo scale