Welcome to SKSurrogate’s documentation!

Contents:

• Introduction
  • Reliability features
  • Dependencies
  • Download
  • Installation
  • Documentation
  • License
    • MIT License
• Surrogate-Based Optimization
  • Sampling
    • CompactSample
    • BoxSample
    • SphereSample
  • Surrogate models
  • Optimizer
• Hyperparameter Optimization
• Hilbert Spaces
  • Orthonormal system of functions
• Sensitivity Analysis
  • Morris
  • Sobol
  • Moment-Independent \(\delta\) Index
• Eliminate features based on Pearson correlation
• Evolutionary Optimization Algorithms
  • Algorithm
  • Quick start
  • Fitness callback contract
  • Constructor parameters
  • Custom operators
  • Pipeline populations
  • Checkpointing and resuming
• Optimized Pipeline Detector
  • Some Technical Notes
    • Stacking
    • Permutation Importance
    • imblearn pipelines
    • Categorical Variables
• Automated Data Type Recognition
  • Example:
• A machine learning progress tracker
  • Runnable example
    • SVM Classifier with RBF v.s. SGDClassifier with Kernels
• Model Bundles and Registry
  • Creating and loading a bundle
  • Registry lifecycle
  • Artifact retention and deletion
  • MLflow-compatible export
  • Compatibility metadata
• Batch Inference
  • Input validation
  • Command-line batch prediction
  • HTTP serving
• Drift Monitoring
  • Report sections
  • Governance and fairness checks
• Web UI and Control-Plane API
  • Local setup
  • Configuration and storage
  • API and UI coverage
• Step-by-Step Web UI User Manual
  • Before you start
  • The workflow at a glance
    • 1. Open the Dashboard and select a task
    • 2. Register and validate datasets
    • 3. Inspect and reduce features (optional)
    • 4. Search model pipelines
    • 5. Evaluate candidate models
    • 6. Create and inspect model bundles
    • 7. Run quality gates before release
    • 8. Register, approve, and promote a model
    • 9. Run online or batch inference
    • 10. Monitor serving quality and data drift
    • 11. Configure a retraining run
    • 12. Follow jobs and audit activity
  • Feature-to-library reference
• Phase 9: Execution and CI/CD Operations
  • Execution backends
    • Local durable execution
    • Dask execution
  • CI model quality gates
  • Triggered retraining
  • Approval-gated deployment and rollback
  • Operational use cases
  • Safety and boundaries
• Phase 10: Documentation and Release Hardening
• Migrating Existing State
• API Reference
• Performance Benchmarks
• Supported Runtime Policy
• Release Checklist and Compatibility Matrix
  • Release checklist
  • Compatibility matrix
• Code Documentation
  • Surrogate Random Search
  • BaseSample
    • BaseSample.check_constraints()
  • BoxSample
    • BoxSample.sample()
  • Categorical
  • CompactSample
    • CompactSample.sample()
  • HDReal
  • Integer
  • Real
  • SphereSample
    • SphereSample.sample()
  • SurrogateRandomCV
    • SurrogateRandomCV.fit()
    • SurrogateRandomCV.pareto_frontier()
  • SurrogateSearch
    • SurrogateSearch.progress()
  • Hilbert Space based regression
  • Error
  • FunctionBasis
    • FunctionBasis.Fourier()
    • FunctionBasis.Poly()
  • FunctionSpace
    • FunctionSpace.FormBasis()
    • FunctionSpace.Series()
    • FunctionSpace.inner()
    • FunctionSpace.project()
  • HilbertRegressor
    • HilbertRegressor.fit()
    • HilbertRegressor.predict()
  • Measure
    • Measure.integral()
    • Measure.norm()
  • Regression
    • Regression.SetFuncSpc()
    • Regression.SetMeasure()
    • Regression.fit()
  • Sensitivity Analysis
  • CorrelationThreshold
    • CorrelationThreshold.fit()
    • CorrelationThreshold.fit_transform()
  • SensAprx
    • SensAprx.fit()
    • SensAprx.fit_transform()

See also the Web UI & API design plan.

Indices and tables

• Index

• Module Index

• Search Page