🦈 Shark Foraging Prediction

This project predicts foraging hotspots of sharks in the North Atlantic using machine learning and geospatial analysis.

📌 Overview

• Preprocessed satellite and environmental data (SST, chlorophyll, depth, etc.).
• Trained a LightGBM model to predict shark foraging probability.
• Generated a geographic grid and filtered out land points using GeoPandas.
• Created an interactive Leaflet.js heatmap to visualize predicted foraging hotspots.

🚀 Features

• Machine learning model for probability estimation
• GeoJSON export (hotspots.geojson)
• Interactive heatmap visualization

⚙️ Tech Stack

• Python (pandas, geopandas, lightgbm)
• Leaflet.js for visualization
• Natural Earth shapefiles for coastline filtering

📂 Outputs

• hotspots.geojson – model predictions as geospatial data
• Interactive heatmap of predicted foraging areas