UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

Is the noise still going on? Predicting repeat noise complaints with historical time course and random forest classifiers

Fan, Zicheng; Signorelli, Valerio; (2023) Is the noise still going on? Predicting repeat noise complaints with historical time course and random forest classifiers. Environment and Planning B: Urban Analytics and City Science 10.1177/23998083231184254. (In press). Green open access

[thumbnail of EPB-2022-0613.R1_Proof_hi.pdf]
Preview
Text
EPB-2022-0613.R1_Proof_hi.pdf - Other

Download (1MB) | Preview

Abstract

Noise can have serious adverse effects on residents' physical and mental health. Since the COVID-19 pandemic, the City of Westminster in London has seen a continuous increase in noise complaints, with a significant number of repeat complaints from the same address within a short time scale. The authorities' ability to respond to complaints is challenged. This study explores a method for predicting and identifying repeat complaints to improve the efficiency of the authorities in dealing with noise complaints. Taking the noise complaint records of the City of Westminster during 2018–2022 as research objects, the research explores the cumulative distribution characteristics and clustering pattern of noise complaints in different spatial and temporal dimensions. On this basis, for a noise complaint from a specific address, the study fits random forest classifiers to predict whether the same address is likely to have another noise complaint in future time scales. It is found that about 18.5% of all complaints had at least one previous complaint at the same address in the previous 7 days; during the lock-down period caused by the COVID-19 pandemic, areas with active commercial activities and higher housing prices experienced a significant decrease in complaints, while areas adjacent to parks and green spaces can share a similar upward trend in noise complaints. Prediction of repeat noise complaints with random forest classifiers is proved feasible. F1 scores of models to predict repeat complaints within 0 to 2nd days, 0 to 7th days and 0 to 30th days in the future are 0.55, 0.66 and 0.75, respectively. Suggestions are provided for local authorities to improve resource allocation related to noise complaint management.

Type: Article
Title: Is the noise still going on? Predicting repeat noise complaints with historical time course and random forest classifiers
Open access status: An open access version is available from UCL Discovery
DOI: 10.1177/23998083231184254
Publisher version: https://doi.org/10.1177/23998083231184254
Language: English
Additional information: This version is the author accepted manuscript. For information on re-use, please refer to the publisher’s terms and conditions.
Keywords: Noise complaints, time series analysis, spatial and temporal clustering, random forest classifier
UCL classification: UCL
UCL > Provost and Vice Provost Offices > UCL BEAMS
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of the Built Environment
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of the Built Environment > Centre for Advanced Spatial Analysis
URI: https://discovery.ucl.ac.uk/id/eprint/10173112
Downloads since deposit
30Downloads
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item