Advanced Experimentation

Add to your intelligent experimentation with SigOpt’s advanced offerings, and customize your optimization workflow by applying expertise of the root problem to enhance our search.

This section will cover the various features available to help you frame intelligent experimentation by customizing the setup of a SigOpt Experiment.

Advanced Features

• Multimetric Optimization - find the Pareto optimal tradeoff between two competing objective metrics.

• Metric Thresholds - restrict the solutions to a subset of the Pareto frontier.

• Metric Constraints - pose auxiliary metrics as constraints.

• Metric Failure - mark evaluation as a failure.

• All Constraints Experiment - search for diverse solutions in more than two metrics.

• Multisolution - search for multiple high performing solutions for a single metric.

• Parallelism - run a SigOpt Experiment on multiple machines.

• Prior Beliefs - define the distribution of a parameter.

• Parameter Constraints - restrict the parameter space in with linear inequality constraints.