Surrogate model construction

Approximate an expensive simulator using a traceable design of experiments.

design samples→run simulation batch Arun simulation batch N→assemble dataset→train surrogate→validate against simulator→publish model

Surrogate model construction

Activity responsibilities

Activity	Responsibility
design-samples	Execute the design samples stage and publish its declared outputs for downstream activities.
run-simulation-batch-a	Execute the run simulation batch A stage and publish its declared outputs for downstream activities.
run-simulation-batch-n	Execute the run simulation batch N stage and publish its declared outputs for downstream activities.
assemble-dataset	Execute the assemble dataset stage and publish its declared outputs for downstream activities.
train-surrogate	Execute the train surrogate stage and publish its declared outputs for downstream activities.
validate-against-simulator	Execute the validate against simulator stage and publish its declared outputs for downstream activities.
publish-model	Execute the publish model stage and publish its declared outputs for downstream activities.

Inputs

• Simulator
• Parameter space
• Sampling policy

Outputs

• Simulation dataset
• Surrogate
• Validity report

Provenance to preserve

• Simulator version
• Samples
• Error by region

Runnable example

This is the complete checked-in bundle for this pattern. Download the environment, scope, topology, workflow, input, container recipe, runner, and validation contract from this page before executing it.

Read setup instructions ↓Environment YAML ↓Execution scope YAML ↓Network topology YAML ↓Workflow YAML ↓Input data ↓Container recipe ↓Run the workflow ↓Validate outputs ↓Expected output manifest ↓

Verified local execution

These captures and the output manifest were produced by the fixture's local Docker run and validator. They are published with the same bundle as the runnable files.

Workflow evidence

Execution evidence

Output evidence

Open verified output manifest ↓

Execution considerations

Run simulations as HPC jobs and training on suitable accelerators. Preserve failed simulations and validity boundaries.

AkôFlow boundary: the engine schedules, deploys, executes, transfers data, and records evidence. The ML or agent framework remains an implementation choice inside each activity.