Automated classification of urban locations for environmental noise impact assessment on the basis of road-traffic content

作者：

Highlights：

• Sampled urban locations are grouped into 4 clusters according to HV and MM content.

• SMO-models outperform MLP-models in correctly classifying obtained clusters.

• Attribute evaluation algorithms achieve better results than subset evaluation algorithms.

• Environment variables, LF sound levels and Leq are the most influential input variables.

摘要

•Sampled urban locations are grouped into 4 clusters according to HV and MM content.•SMO-models outperform MLP-models in correctly classifying obtained clusters.•Attribute evaluation algorithms achieve better results than subset evaluation algorithms.•Environment variables, LF sound levels and Leq are the most influential input variables.

论文关键词：Machine-learning,Feature selection,Classification,Road traffic noise,Noise impact,Urban environments

论文评审过程：Received 13 May 2015, Revised 10 January 2016, Accepted 11 January 2016, Available online 15 January 2016, Version of Record 5 February 2016.

论文官网地址：https://doi.org/10.1016/j.eswa.2016.01.011