Optimization components#
This section organizes the components that make up an optimization, including Component, Problem, Population, and Comparator. The pages grouped by responsibility provide details about the built-in components. See Framework for how contracts become execution plans.
Problem definition, constraint handling, ranking, and non-dominated solutions.
Candidate generation, mutation, parent selection, and survivor selection.
Training data, prediction, acquisition scoring, and model switching.
Initialization, candidate evaluation, generation processing, and termination.
Event observation, the Stage compatibility boundary, execution state, and population data.
How to choose among swapping components, hooks, Stages, Callbacks, and framework extensions.
For your first optimization, see the Quickstart; to choose an extension point, see the Extension guidelines. See the API reference for public import paths for each type.