Welcome to IMPROVE’s documentation!

Warning

This documentation is for a previous release of IMPROVE (v0.0.3-beta). You can see the latest release of IMPROVE here and the latest version of IMPROVE here.

The IMPROVE project aims to standardize execution of supervised learning models to facilitate benchmarking and comparison of models. We have created a Python-based IMPROVE framework that encases community models and IMPROVE benchmarking schemes for model comparision. As a seminal case, we have curated community models and a benchmark dataset for single drug response prediction. Do you have a model you would like to easily compare to state-of-the-art models? Try IMPROVE!

Contents:

• What is the IMPROVE project?
  • Background
  • Future Directions
  • Acknowledgments
• Quickstart
• Benchmark Evaluation Schemes
  • Cross-Study Analysis
  • Learning Curve Analysis
• Drug Response Prediction
  • Problem Formulation
  • Benchmark Data
  • Models
• User Guide
  • Building Containers
  • CSA
• Model Contributor Guide
  • Tutorial

Access the code base https://github.com/JDACS4C-IMPROVE/