Welcome to Renormalized Mutual Information for Artificial Scientific Discovery’s documentation!

Introduction - Reference to Paper

Quick start

Authors

Leopoldo Sarra, Florian Marquardt

Contents:

• Quick Start
  • Estimate Renormalized Mutual Information of a feature
    • Simplest feature
    • Reparametrization invariance
    • Another feature
  • Feature Optimization
  • Tradeoff between simplicity and compression
  • Conclusions
• Estimation
• Feature Extraction
• Examples
  • Spiral-shaped distribution
  • Wave Packet
  • Liquid Drop
  • Examples of one-dimensional features
• Implementations
  • Spiral-shaped distribution
    • Feature Extraction
    • Plot the feature
    • Calculate RMI of the extracted feature
    • Comparison with Principal Component Analysis
  • Wave Packet
    • Neural Network feature extraction
    • Autoencoder
    • Compare different features
    • Supervised task starting from the feature representation
  • Liquid Drop
    • Load a dataset
    • Feature comparison
    • Analyze deformation range dependence
    • Supervised task starting from the feature representation

Indices and tables

• Index

• Module Index

• Search Page