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

In traditional deep learning algorithms, one of the key assumptions is thatthe data distribution remains constant during both training and deployment.However, this assumption becomes problematic when faced withOut-of-Distribution periods, such as the COVID-19 lockdowns, where the datadistribution significantly deviates from what the model has seen duringtraining. This paper employs a two-fold strategy: utilizing continual learningtechniques to update models with new data and harnessing human mobility datacollected from privacy-preserving pedestrian counters located outsidebuildings. In contrast to online learning, which suffers from 'catastrophicforgetting' as newly acquired knowledge often erases prior information,continual learning offers a holistic approach by preserving past insights whileintegrating new data. This research applies FSNet, a powerful continuallearning algorithm, to real-world data from 13 building complexes in Melbourne,Australia, a city which had the second longest total lockdown duration globallyduring the pandemic. Results underscore the crucial role of continual learningin accurate energy forecasting, particularly during Out-of-Distributionperiods. Secondary data such as mobility and temperature provided ancillarysupport to the primary forecasting model. More importantly, while traditionalmethods struggled to adapt during lockdowns, models featuring at least onlinelearning demonstrated resilience, with lockdown periods posing fewer challengesonce armed with adaptive learning techniques. This study contributes valuablemethodologies and insights to the ongoing effort to improve energy loadforecasting during future Out-of-Distribution periods.

cornell university

Navigating Out-of-Distribution Electricity Load Forecasting during  COVID-19: Benchmarking energy load forecasting models without and with  continual learning