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-technical losses (NTL) occur during the distribution of electricity inpower grids and include, but are not limited to, electricity theft and faultymeters. In emerging countries, they may range up to 40% of the totalelectricity distributed. In order to detect NTLs, machine learning methods areused that learn irregular consumption patterns from customer data andinspection results. The Big Data paradigm followed in modern machine learningreflects the desire of deriving better conclusions from simply analyzing moredata, without the necessity of looking at theory and models. However, thesample of inspected customers may be biased, i.e. it does not represent thepopulation of all customers. As a consequence, machine learning models trainedon these inspection results are biased as well and therefore lead to unreliablepredictions of whether customers cause NTL or not. In machine learning, thisissue is called covariate shift and has not been addressed in the literature onNTL detection yet. In this work, we present a novel framework for quantifyingand visualizing covariate shift. We apply it to a commercial data set fromBrazil that consists of 3.6M customers and 820K inspection results. We showthat some features have a stronger covariate shift than others, makingpredictions less reliable. In particular, previous inspections were focused oncertain neighborhoods or customer classes and that they were not sufficientlyspread among the population of customers. This framework is about to bedeployed in a commercial product for NTL detection.

Is Big Data Sufficient for a Reliable Detection of Non-Technical Losses?