This repository contains a Jupyter notebook, the code for a web app and a Dockerfile.
The Jupyter notebook introduces the dataset and contains the modeling steps for binary a classification problem. I compare a logistic regression, random forest and XGBoost classifier.
The web app consists of a Flask backend and a Javascript frontend that serve predictions from the pickled random forest model.
The frontend looks like this:
The API will return a prediction:
Also included is a Dockerfile to download the required libraries and run the web app.
For example, to build the docker image run:
docker build -t no-show .
To create a container witht the web app:
docker run -it -p 8080:8080 no-show
The app will be available at http://localhost:8080/
The data is downloaded from kaggle, from Medical Appointment No Show case.