temporai 0.0.1.dev0

TemporAI: ML-centric Toolkit for Medical Time Series

TemporAI

⚠️ Status: Work in progress, please come back at a later date. We are migrating from a previous iteration of the project.

TemporAI is a Machine Learning-centric time-series library for medicine. The tasks that are currently of focus in TemporAI are: time-series prediction, time-to-event (a.k.a. survival) analysis with time-series data, and counterfactual inference (i.e. [individualized] treatment effects). The library also aims to provide the user with understanding of their data, model, and problem, through e.g. integration with interpretability methods.

Key concepts:

Installation

PiPy release is coming soon, for now install directly from the repository as follows:

pip install git+https://github.com/vanderschaarlab/temporai.git

To view the list of dependencies, see here.

TemporAI Pipeline

The diagram below illustrates the structure of a TemporAI pipeline: