Wildfire Prediction and Heatmap visualization This project aims to predict the likelihood of wildfires based on the satellite features provided and generate a heatmap to visualize potential wildfire locations. This project uses the Random Forest Classifier to build the prediction model and Folium to create a geographical heatmap of wildfire predictions.

Data

The dataset used in this project is sourced from MODIS satellite data in South Africa for the year 2023. The dataset includes latitude, longitude, brightness, scan, track, and fire radiative power (FRP).

Key Features

latitude: Geographic coordinate longitude: Geographic coordinate brightness: Brightness of the fire as captured by the satellite scan and track: Satellite tracking data bright_t31: Temperature at band 31 of the satellite FRP: Fire Radiative Power, the energy emitted by the fire

To run this project, you will need the following Python libraries:

pandas

NumPy

sci-kit-learn

seaborn

matplotlib

folium

Model Metrics:

Accuracy: Evaluates the overall performance of the model. Classification Report: Precision, recall, and F1-score for both classes. Confusion Matrix: Breakdown of true positives, false positives, true negatives, and false negatives.

Results Accuracy: The accuracy of the model on the test data. Heatmap: The generated heatmap shows regions where wildfires are predicted to occur based on the satellite data.