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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Thermal preferences vary from person to person and may change over time. Themain objective of this paper is to sequentially pose intelligent queries tooccupants in order to optimally learn the indoor air temperature values whichmaximize their satisfaction. Our central hypothesis is that an occupant'spreference relation over indoor air temperature can be described using a scalarfunction of these temperatures, which we call the "occupant's thermal utilityfunction". Information about an occupant's preference over these temperaturesis available to us through their response to thermal preference queries :"prefer warmer," "prefer cooler" and "satisfied" which we interpret asstatements about the derivative of their utility function, i.e. the utilityfunction is "increasing", "decreasing" and "constant" respectively. We modelthis hidden utility function using a Gaussian process prior with built-inunimodality constraint, i.e., the utility function has a unique maximum, and wetrain this model using Bayesian inference. This permits an expected improvementbased selection of next preference query to pose to the occupant, which takesinto account both exploration (sampling from areas of high uncertainty) andexploitation (sampling from areas which are likely to offer an improvement overcurrent best observation). We use this framework to sequentially designexperiments and illustrate its benefits by showing that it requires drasticallyfewer observations to learn the maximally preferred temperature values ascompared to other methods. This framework is an important step towards thedevelopment of intelligent HVAC systems which would be able to respond tooccupants' personalized thermal comfort needs. In order to encourage the use ofour PE framework and ensure reproducibility in results, we publish animplementation of our work named GPPrefElicit as an open-source package inPython.

Learning Personalized Thermal Preferences via Bayesian Active Learning with Unimodality Constraints