Sequential Preference-Based Optimization
This work addresses the need for efficient optimization in engineering problems that rely on human feedback, but it is incremental as it builds on prior methods.
The authors tackled the problem of optimizing designs based on human preferences by developing PrefOpt, an open-source package that extends an existing latent variable model to handle equivalent preferences, resulting in a tool for sequential optimization tasks.
Many real-world engineering problems rely on human preferences to guide their design and optimization. We present PrefOpt, an open source package to simplify sequential optimization tasks that incorporate human preference feedback. Our approach extends an existing latent variable model for binary preferences to allow for observations of equivalent preference from users.