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

Non-Intrusive Load Monitoring (NILM) is a field of research focused onsegregating constituent electrical loads in a system based only on theiraggregated signal. Significant computational resources and research time arespent training models, often using as much data as possible, perhaps driven bythe preconception that more data equates to more accurate models and betterperforming algorithms. When has enough prior training been done? When has aNILM algorithm encountered new, unseen data? This work applies the notion ofBayesian surprise to answer these questions which are important for bothsupervised and unsupervised algorithms. We quantify the degree of surprisebetween the predictive distribution (termed postdictive surprise), as well asthe transitional probabilities (termed transitional surprise), before and aftera window of observations. We compare the performance of several benchmark NILMalgorithms supported by NILMTK, in order to establish a useful threshold on thetwo combined measures of surprise. We validate the use of transitional surpriseby exploring the performance of a popular Hidden Markov Model as a function ofsurprise threshold. Finally, we explore the use of a surprise threshold as aregularization technique to avoid overfitting in cross-dataset performance.Although the generality of the specific surprise threshold discussed herein maybe suspect without further testing, this work provides clear evidence that apoint of diminishing returns of model performance with respect to dataset sizeexists. This has implications for future model development, datasetacquisition, as well as aiding in model flexibility during deployment.

Exploring Bayesian Surprise to Prevent Overfitting and to Predict Model Performance in Non-Intrusive Load Monitoring