Getting started

Welcome to the documentation for the No-Show prediction model. This project aims to prevent no-shows by calling high-risk patients before their appointments.

Overview

On this page, you will find information on how to install and use the No-Show prediction model. For information on the dataset used for training the model, please refer to the Dataset Card.

The idea of a no-show prediction model was first implemented at the Erasmus MC, this project is based on their approach and features.

Install the No-Show package

To install the No-Show package, first clone the repository and install the package using a package manager like pip or uv.

clone the repository:

git clone https://github.com/UMCU-Digital-Health/No_Show.git
cd No_Show

Then, install the package:

pip install -r requirements.txt

or using uv:

uv sync

Run pipelines

To run the entire pipeline from data export to model training, you can use the train_no_show command (or python src/noshow/train_pipeline.py):

train_no_show --skip-export  # skip the export step if you already have the data

For more information on data used, check the dataset card here

Run the API

To run the prediction API, which creates predictions for appointments, you can use the following command:

python run/app.py

Run the Calling Dashboard

The calling dashboard is a Streamlit application that allows users to view predictions and manage patient calls. To run the dashboard locally, use the following command:

streamlit run run/calling_dash.py

Run the Admin dashboard

The admin dashboard is a Streamlit application that allows users to check the performance of the application and call results. To run the dashboard locally, use the following command:

streamlit run run/admin_dash.py