Models to Predict Pandemic Spread

model implementation in python

Inspiration

In epidemeology, there are certain mathematical models to predict disease spread. These models are compartmental, meaning they divide a model population into different groups (usually Susceptible, Infected, and Recovered). No library or tool exists which lets users to add nodes/compartments to account for variables they might want to study.

What it does

Provides an accessible UI/UX to implement compartmentalized models and corresponding flow differential equations.

How we built it

We did research into how the basic mathematical model works. Then, we built a Blocks class which allows users to input compartments and route their inputs and outputs to other compartments and define relationships (linear or non-linear). The Blocks are then parsed by a Model class which also creates the differential equations and solves them using Euler's Step method (numerical integration).

Challenges we ran into

Finding relevant packages to implement frontend and backend apps
Parsing and cleaning data
Finding relevant data for model
Implementing ML techniques ## Accomplishments that we're proud of
Functional Differentiation Equation Solver ## What we learned
How to manage time efficiently to merge together parallel workflows ## What's next for Models to Predict Pandemic Spread
Polishing the backend model to include more non-linear relationships
Polishing the front end
Incorporating ML techniques to better fit real world data and to make a generative disease spread model.