Perpetual Documentation

Perpetual is a self-generalizing gradient boosting machine that doesn’t need hyperparameter optimization. It is designed to be easy to use while providing state-of-the-art predictive performance.

Key Features

• Hyperparameter-Free Learning: Achieves optimal accuracy in a single run via a simple budget parameter, eliminating the need for time-consuming hyperparameter optimization.

• High-Performance Rust Core: Blazing-fast training and inference with a native Rust core, zero-copy support for Polars/Arrow data, and robust Python & R bindings.

• Comprehensive Objectives: Fully supports Classification (Binary & Multi-class), Regression, and Ranking tasks.

• Advanced Tree Features: Natively handles categorical variables, learnable missing value splits, monotonic constraints, and feature interaction constraints.

• Built-in Causal ML: Out-of-the-box support for causal machine learning to estimate treatment effects.

• Robust Drift Monitoring: Built-in capabilities to monitor both data drift and concept drift without requiring ground truth labels or model retraining.

• Continual Learning: Built-in continual learning capabilities that significantly reduce computational time from O(n²) to O(n).

• Native Calibration: Built-in calibration features to predict fully calibrated distributions (marginal coverage) and conditional coverage without retraining.

• Explainability: Easily interpret model decisions using built-in feature importance, partial dependence plots, and Shapley (SHAP) values.

• Production Ready & Interoperable: Ready for production applications; seamlessly export models to industry-standard XGBoost or ONNX formats for straightforward deployment.

Contents:

• Installation
  • Requirements
  • Optional Dependencies
• Quickstart
  • Classification
  • Regression
  • Ranking
  • More Examples
• API Reference
  • PerpetualBooster
  • Sklearn Interface
  • Causal ML
  • Double Machine Learning
  • Uplift Modeling
  • Policy Learning
  • Causal Metrics
  • Fairness
  • Regulatory Risk
• Causal ML
  • Uplift Modeling
  • Double Machine Learning (DML)
  • Policy Learning
  • Instrumental Variables (BoostIV)
  • Regulatory Risk & Interpretability
  • Fairness and Algorithmic Bias
• Calibration and Uncertainty Quantification
  • The Fundamental Advantage: Post-Hoc Calibration
  • Probability Calibration (Classification)
  • Uncertainty Quantification (Regression)
  • Raw Prediction Distributions
  • Why Perpetual is Better
  • Tutorials
  • Next Steps
• Drift Detection
  • How it Works
  • Usage
  • Interpreting the Score
  • Examples
• Continual Learning
  • Mechanism
  • Warm Start vs Retraining
  • Performance Considerations
  • Example
• Explainability
  • Feature Importance
  • Partial Dependence
  • Prediction Contributions
  • Example
• Model IO & Export
  • Native Serialization
  • Pickling Support
  • Metadata Persistence
  • Exporting to XGBoost
  • Exporting to ONNX
  • Example
• Tutorials
  • Handling Categorical Data
  • Scikit-Learn Interface: Classification, Regression & Ranking
  • Performance Benchmarking
  • Mastering Regression Calibration: From Theoretical Basics to Advanced Methods
  • Classification Calibration: Prediction Sets with Perpetual
  • Custom Objective Tutorial
  • Uplift Modeling with the Criteo Uplift Dataset
  • Instrumental Variables (Boosted IV)
  • Double Machine Learning: Estimating the Gender Wage Gap
  • Policy Learning: Optimal Treatment Assignment
  • Risk, Compliance, and Interpretability
  • Heterogeneous Treatment Effects with Meta-Learners
  • Fairness-Aware Credit Scoring
  • Fairness-Aware Classification with FairClassifier
  • Customer Retention: Uplift Modeling for Churn Prevention
  • Advanced Drift Detection in Perpetual
• Architecture
  • System Overview
  • Rust Core
  • Python Interface
  • PerpetualBooster Algorithm
• Parameters Tuning
  • The Budget Parameter
  • How it works
• Frequently Asked Questions
  • How do I choose the best budget?
  • Does Perpetual support categorical features?
  • Is Perpetual faster than XGBoost?

API Reference

See the API Reference for the detailed API reference.

Indices and tables

• Index

• Module Index

• Search Page