Inspiration
We are all interested in using data science to model economic processes, and felt like hot dogs was a very interesting problem because it is like a microcosm of the larger US economy.
What it does
Our model takes in as input a variety of attributes of a store (ie. if it has a car wash, serves alcohol) as well as a day of the year and time, and uses these to predict the gross hot dog sold in that given time on the given day of the year.
How we built it
We used a Random Forest Regressor as well as an 80-20 train-test split to create our model. We decided on a Random Forest Regressor because we felt like it best captured the combination of continuous and categorical variables in our model.
Challenges we ran into
We had to try a variety of different models (neural networks, time series analysis, etc), and it took a couple different tries to find a model that seemed to work well for the data.
Accomplishments that we're proud of
This was our first delve into the world of data science and data analysis, and we learned a lot of skills in pandas and scikit-learn along the way.
What we learned
Pandas, scikit-learn, dealing with categorical variables
What's next for Hotdawg Sale Prediction
Would like to try and see if this model is adaptable to other consumer goods sold in gas stations
Log in or sign up for Devpost to join the conversation.