Use of Machine Learning in the Analysis of IndoorELF MF Exposure in Children

Characterization of children exposure to extremely low frequency (ELF) magnetic fieldsis an important issue because of the possible correlation of leukemia onset with ELF exposure.Cluster analysis--a Machine Learning approach--was applied on personal exposure measurementsfrom 977 children in France to characterize real-life ELF exposure scenarios. Electric networks nearthe child's home or school were considered as environmental factors characterizing the exposurescenarios. The following clusters were identified: children with the highest exposure living 120-200 mfrom 225 kV/400 kV overhead lines; children with mid-to-high exposure living 70-100 m from63 kV/150 kV overhead lines; children with mid-to-low exposure living 40 m from 400 V/20 kVsubstations and underground networks; children with the lowest exposure and the lowest number ofelectric networks in the vicinity. 63-225 kV underground networks within 20 m and 400 V/20 kVoverhead lines within 40 m played a marginal role in differentiating exposure clusters. Clusteranalysis is a viable approach to discovering variables best characterizing the exposure scenarios andthus it might be potentially useful to better tailor epidemiological studies. The present study did notassess the impact of indoor sources of exposure, which should be addressed in a further study.