A system for determining noise-laden regions using data mining approaches and cluster techniques

摘要

A system for determining the regions affected by noise pollution, comprising:a data collection unit for recording different regions of a site affected by noise, both day and night, and for storing in a data set;a preprocessing unit to remove missing/noisy values from the dataset and calculate the maximum normalization, which converts the data to the normalized form for each year and summarizes the data in an Excel table;a cluster-wide data mining approach to assess the main causes of noise pollution in the clustered areas, clustering the regions according to the level of noise pollution in the area concerned; anda classification unit comprises a decision tree consisting of root nodes and leaf and intermediate nodes in order to identify the main and predominant reasons responsible for the occurrence of noise pollution in the regions in a given year using a classification procedure; where the dataset is classified on the basis of a cluster analysis.