EcoBite

Inspiration

We did preliminary research on the leading causes of food waste and felt like our app could increase communication between different channels to help optimize food waste for consumers, farmers, and food service companies.

What it does

Our implementations bridge the gap between farms, food companies, and consumers by providing AI fueled insights.

How we built it

For the Random Forest Regressor, we used Python in JupyterNotebook. We collected the dataset from Kaggle, which shows agriculture product sales data. To help this implementation, I referred to resources from a past class that reviewed Random Forest Regressor and used ChatGPT. Once I implemented the model, I calculated relevant errors and created data visualizations.

For Figma, we built a simple website demo from the farmers account perspective and an app from the consumers profile. We used ChatGPT for text content and outline idea generation.

Challenges we ran into

Before landing on the Random Forest Regressor, I struggled to work with ARIMA and SARIMA models on a dataset that showed rice sales in Pakistan. The predictions were not accurate and because there was a downward trend and no restriction on negative values, the predictions went negative over a long-term period, which is not helpful. I had implemented a Random Forest Regressor before in a class and decided to use a different dataset.

Accomplishments that we're proud of

We spent a lot of time working on different models and got a hands-on sense of how the models work, and more importantly, why they didn't work with my datasets. We are proud of the final result that came together. Designing the app and website was tedious as we paid a lot of attention to details. We made the app interactive created a project we are happy with.