COVID-19 Food Fight

Inspiration

Amidst the COVID-19 pandemic, which affects millions of people worldwide and has caused hundreds of thousands of deaths, many factors determine whether or not you are at risk of infection. We focused on Health and Wellness and the relationship between diet and coronavirus Unfortunately, risk factors are difficult to pinpoint because of the complicated nature of the virus. Underlying health conditions muddle scientists’ and doctors’ abilities to fully understand the pandemic. A contributor is diet. Different countries also have unique cuisines and diets that need to be accounted for. While researching about coronavirus, we came across a Kaggle dataset that compiled food intake, nourishment values, and coronavirus statistics. This dataset stood out to us because of its clean compiled nature. We were also very interested about the relationship diet may have with Coronavirus because it was something so new to us.

What it does

We allow users to select the foods they eat and then we tell them how at risk they are based on what the machine learning algorithm predicts. We classified foods by their coefficient of correlation to confirmed cases in their area then created a threshold based on the highest and lowest values. If the food had a coefficient of correlation lower than the threshold then it was low risk and if it had a coefficient of correlation higher than the threshold it was high risk. The website would then score the user's inputs and tell them if they were high or low risk. The website would also show articles that could be useful for more research.

How we built it

We have a background in machine learning, and know that ML is a useful tool for making predictions. We used python and JupyterNotebook to create the machine learning algorithm with linear regression. In order to make our predictions and discoveries comprehensive we decided to build a HMTL driven webpage on the frontend. Then, to get the backend and front end to communicate we used a little bit of JavaScript.

Challenges we ran into

The data set needs to be more comprehensive and we could improve by integrating front end and back end.

Accomplishments that we're proud of

Our project uniquely uses innovative machine learning. Analyzing diet offers a new perspective to the current conversation on coronavirus. We think our project’s inventive methods and unique subject matter puts us apart from other projects.

What we learned

We learned a lot about web development especially how to create "business" layer that communicates between front end and backend. We also learned how to research and where to find good datasets for COVID-19.

What's next for COVID-19 Food Fight

We would like to improve the machine learning algorithm to predict risk accurately, preferably by classifying diets into low or high risk. We would also like to add a feature that gives recommendations on foods depending on someone’s specific diet in order to reduce risk. We can also add features that factor in other lifestyle choices like exercise and sleep schedule We would like to market the website through partnering with larger fitness apps such as MyFitnessPal which also tracks your diet.