Expected improvement
Expected improvement is a Bayesian optimization acquisition function that estimates a candidate’s expected gain over the best observed objective value while accounting for predictive uncertainty.
Bayesian Optimization in Action
Expected improvement is a Bayesian optimization acquisition function that quantifies the expected amount by which evaluating a candidate may improve on the best observed objective value.
- How it is used
- Ranks candidate evaluations in Bayesian optimization by expected improvement over the incumbent.
- How deep it goes
- Develops EI from improvement intuition through probabilistic magnitude, closed-form scoring, BoTorch use, and a batch extension.
- Examples
- Uses validation-accuracy tuning, a one-dimensional objective search, and BoTorch acquisition construction.
- What you can do after
- Apply it via BoTorch to select and optimize the next Bayesian optimization query.
Read how Bayesian Optimization in Action explains expected improvement →
AI-Powered Search
Expected improvement is a Bayesian-optimization acquisition function that scores candidates by their expected gain over the current best result while accounting for predictive uncertainty.
- How it is used
- Guides active-learning exploration by choosing search candidates with the highest predicted upside for live-user evaluation.
- How deep it goes
- Explains the basic formula and implementation, but defers advanced Bayesian-optimization details.
- Examples
- Uses product features, logged judgments, uncertainty tables, and a "transformers dvd" exploration query.
- What you can do after
- Apply it via Python to rank candidate feature vectors and select products for search-engine exploration.
Related concepts
- constrained expected improvementincludes · Constrained expected improvement is a Bayesian optimization acquisition function whose score is the expected improvement multiplied by the probability that the candidate satisfies the constraints.
- ExpectedHypervolumeImprovementincludes · ExpectedHypervolumeImprovement is a BoTorch acquisition-function class that computes analytic expected hypervolume improvement for multiobjective Bayesian optimization.
- Bayesian optimizationkind of · Bayesian optimization is a sequential method for optimizing an expensive or unknown objective by using a probabilistic model to guide evaluations.
- policykind of · A policy is a rule, model, or set of rules that determines behavior, decisions, or permitted actions under specified conditions.
- acquisition functionkind of · An acquisition function is a scoring rule that ranks candidate evaluations by their expected usefulness for selecting the next sample or experiment.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.