Welcome to the PermutationImportance documentation!

PermutationImportance is a Python package for Python 2.7 and 3.5+ which provides several methods for computing data-based predictor importance. The methods implemented are model-agnostic and can be used for any machine learning model in many stages of development. For more information on the particular methods, please see the documentation for that particular method.

Contents:

• About PermutationImportance
• Installation
• Levels of Abstraction
  • Model-Based
  • Method-Specific
  • Method-Agnostic
• Methods
  • Permutation Importance
    • Theory
    • Usage
    • References
  • Sequential Selection
    • Theory
    • Usage
    • References
  • Custom Methods
    • Abstract Variable Importance
    • Selection Strategy
    • Example of Custom Variable Importance
• License
  • MIT License
• Modules
  • Abstract Runner
  • Data Verification
  • Error Handling
  • Metrics
  • Multiprocessing Utils
  • Permutation Importance
  • Result
  • Scoring Strategies
  • Selection Strategies
  • Sequential Selection
  • Sklearn API
  • Utils