Wildfire-Detection-AI_Project-

Wildfire Risk Prediction in Canada Using Satellite Imagery Overview Using deep learning and satellite images, this project aims to predict wildfire occurrences in Canada to assist in early warning and resource allocation. Research Question: Can deep learning models predict wildfires in Canada from satellite imagery? Which image features are most significant? Dataset: Wildfire Prediction Dataset: 42,850 satellite images from Kaggle. Data Split: Training (70%), Validation (15%), Test (15%). Methodology Model: CNNs for image classification. Bias Considerations Data Imbalance: Unequal class distribution. Temporal & Geographical Bias: Seasonal and regional limitations. Goals Predict wildfire occurrences. Provide visual insights for model decisions. Support preventive measures and resource allocation.