Inspiration
The increased intensity and risk of Wildfires are a pressing concern as global temperatures are rising and contributing to a change in overall climate around the world. It not only has economic impacts, but social and health impacts to inhabitants of the region, especially indigenous people.
What it does
We have designed analysis and models for Alberta Wildfires prediction for the business use case of helping government leaders, firefighters and Wildfire Management employees reduce the extreme costs and health risks associated with wildfires.
How we built it
We used combination of python data analysis tools such as pandas, matplotlib for doing data analysis and used classification models from sklearn such as XGBoost classifiers for predicting wildfire severity and size
Challenges we ran into
We found a few models from sklearn used for classification. However, finding more different architectures of machine learning models would be something that we would look into. Also, finding the features that contributed to an increased severity and risk of wildfires we had to do multiple plots through our research and discuss amongst each other to find true causes.
Accomplishments that we're proud of
We dug into the data to understand what were the insights gleaned from the data. We also did a lot of research into understanding how wildfires impact indigenous populations, and different types of forest environments. This is despite the fact that not much information is available online regarding impacts on indigenous populations. Also, two members had their first time doing a datathon and they contributed a lot to the development of the project.
What we learned
Data Analysis, Model Development, EDA, Data Visualization, Group Work, Group Dynamics
What's next for Him
We plan to look further into how the government is acting on indigenous practices like cultural burns, are indigenous people getting enough resources to support their living in their lands, and if indigenous people can return to their reserves after the wildfire. Also, we want to further look at cost data and how particularly severe and large area wildfires contribute to federal and provincial spending on prevention of wildfires. Look further into the effect of how different regions, different fuel types, and various weather changes affect how long a fire lasts. This will also aid firefighters and government personnel to understand how to fight fires to reduce overall negative impact on costs and people's health.
Log in or sign up for Devpost to join the conversation.