CA-Wildfire-Risk-Prediction-and-Optimization

Wildfire Risk Prediction and Response Optimization in California:

• Exploring Climate Change and Inmate Firefighters
• Utilizing Random Forest, SMOTE, Constraint Optimization, Mixed-Integer Programming

Description of Key Notebooks:

• california_map.ipynb: makes chloropleth maps of California to show wildfire risk predictions
• constaint_data.ipynb: data wrangling and EDA for firefighter numbers, station locations, and inmate firefighter population in California
• data_wrangling.ipynb: data wrangling and EDA for historic wildfire incident, weather, and topographic information used for Tree-based risk prediction model
• full_analysis.ipynb: Tree- based risk prediction models taking advantage of undersampling approaches
• optimization.ipynb: Mixed Integer Programming work for Optimal firefighter allocation