Based on the data hosted by (kaggle)[https://www.kaggle.com/residentmario/ramen-ratings] which is from a popular ramen review site.
The goal of the project is for its contributors to become more comfortable with scikit-learn pipeline, custom pipeline functions, hyperparameters tunning and resulting machine learning model deployment on Django.
The deployed Django model is interactive meaning that a user can enter several parameters including ingredients, country, style and company to have the model return a predicted rating.
Some of the finding so far:
- the data is imbalanced with many 5 stars reviews;