Wildfire Risk Assessment is an innovative project leveraging Canada's open-source forest wildfire data to advance prediction and risk management strategies. This tool is crucial for businesses and organizations involved in environmental safety, forestry management, and emergency response planning.
The project has two primary goals:
- Predict the Cause of Forest Fires: Utilizing machine learning, the project aims to identify whether a fire is caused by human activities or lightning, refining response strategies and preventive measures.
- Predict the Probability of Fire Occurrence: Forecasting the likelihood of fire occurrences based on location and time data, aiding in resource allocation and risk mitigation.
- Risk Management: Enhance wildfire preparedness and response strategies.
- Resource Allocation: Optimize the deployment of firefighting resources.
- Insurance and Finance: Inform risk assessments for insurance and investment purposes.
- Environmental Conservation: Support proactive measures in forest management and conservation efforts.
- Data-Driven Approach: Leverages historical fire data and temperature records.
- Advanced Modeling: Employs machine learning for accurate and reliable predictions.
- User-Friendly Interface: Easy to integrate into existing systems for real-time decision making.
- Data Files:
forest_fire.txtcontains the raw data needed. - Analysis Tools: Access the Jupyter notebook
data_cleaning.ipynbfor data processing insights. - Models Used: Using Random Forrest and Gradient Boosting to generate interpretable results.
- Directories:
Prong1_Predicting_Unknown_Fires: Contains models and scripts for predicting the causes of fires with unknown origins.Prong2_Predicting_Fire_Probability: Houses models and scripts for calculating the probability of fire occurrences at specific locations and times.
Wildfire Risk Assessment - Predicting and managing wildfire risks with data-driven insights.