Purpose
Our project aims to utilize the power of machine learning to provide an accurate and timely
prediction of wildfires as a means of prevention.
One of our goals is to improve the machine learning algorithm.
Another one of our goals is to design a website that shows the results of the machine learning
algorithm and visualizes the data in a user-friendly way.
Features
One of the features our platform offers is 2D visualizations of predictions from our trained
models using algorithms such as, Gaussian Naive Bayes (GNB), K-Nearest Neighbors (KNN), Long
Short-Term Memory (LSTM), Support Vector Machine (SVM), and Multilayer Perceptron (MLP).
Our second feature for our platform offers 3D visualizations of processed data such as enhanced
vegetation index (EVI), land surface temperature (LST), thermal anomalies (TA), burned area
(BA), average wind data (AWS), and fire records (FR). Users will be able to view an animation
of this data based on a wildfire heavy year for California:2020.
Our last feature
allows for user-friendly access to downloadable processed data spanning years 2013-2020,
ensuring accessibility and convenience for anyone in the community.
Technologies,
Frameworks, & Skills
The technologies and frameworks used to create our platform were:
Jupyter: Jupyter provides
acomprehensive,
interactive, and versatile environment that supports the entire machine learning workflow: from
data preprocessing and exploration to modeling, evaluation, and presentation.
Discord: Main form of
communication within the group.
Messaging, Voice Chat, Video Sharing.
Visual Studio:
Project IDE that works well with Jupyter
GitHub: Allows Collaboration and
Keeps track of project through branches.
Jira:
Software that keeps tasks and project issues on track.
Tableau: Tableau is a
powerful data visualization software that allows users to create
interactive and shareable visualizations from various data sources.
Blender :a free and
open-source 3D creation suite that supports the entire 3D pipeline,
including modeling, rigging, animation, simulation, rendering, compositing, and motion tracking.
Flask:a
lightweight and
extensible web framework for Python that allows developers to quickly build web applications
with minimal boilerplate code.
The skills required for this project were:
Machine learning & Python: Machine Learning was
used to process preprocessed data using python code. It was also used to train models to predict
where a fire would occur.
Web design, HTML & CSS: These skills were all used and needed to make a user friendly
website to display visualization of machine learning work.