2026-10-16T00:52:43Z https://data.mendeley.com/oai

oai:data.mendeley.com/7mgw79zynr.1 2026-09-15T09:26:12Z

Singh, Ninni Analysis code and derived results for Measuring the Effect of Instrumentation Coverage on Appliance Disclosure Using Gradient-Boosted Trees across 25 Households Mendeley Data The article asks what a smart meter record reveals about household activity, and shows that the answer depends heavily on something nobody reports: the fraction of a household's load that the dataset actually submeters. Appliances left unmetered still draw power, still reach the aggregate channel, and there they mask the appliances a classifier is trying to detect. Across 25 households that fraction ranges from 0.003 to 0.805, and completing a household's instrumentation raises measured disclosure in every one of them, by amounts that differ from home to home. Computer Science Privacy Gunjan, Vinit Kumar Dataset doi:10.17632/7mgw79zynr.1 oai:data.mendeley.com/7mgw79zynr.1 Creative Commons Attribution 4.0 International http://creativecommons.org/licenses/by/4.0 https://data.mendeley.com/datasets/7mgw79zynr 2026-09-15T09:26:12Z