English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Model-based optimization (MBO) is increasingly applied to design problems inscience and engineering. A common scenario involves using a fixed training setto train models, with the goal of designing new samples that outperform thosepresent in the training data. A major challenge in this setting is distributionshift, where the distributions of training and design samples are different.While some shift is expected, as the goal is to create better designs, thischange can negatively affect model accuracy and subsequently, design quality.Despite the widespread nature of this problem, addressing it demands deepdomain knowledge and artful application. To tackle this issue, we propose astraightforward method for design practitioners that detects distributionshifts. This method trains a binary classifier using knowledge of the unlabeleddesign distribution to separate the training data from the design data. Theclassifier's logit scores are then used as a proxy measure of distributionshift. We validate our method in a real-world application by running offlineMBO and evaluate the effect of distribution shift on design quality. We findthat the intensity of the shift in the design distribution varies based on thenumber of steps taken by the optimization algorithm, and our simple approachcan identify these shifts. This enables users to constrain their search toregions where the model's predictions are reliable, thereby increasing thequality of designs.

Beyond the training set: an intuitive method for detecting distribution shift in model-based optimization