Stochastic Dosimetry and Machine Learning: Innovative Approaches for Facing Challenges in Exposure Assessment in Realistic Scenarios

Innovative approaches, such as stochastic dosimetry and Machine Learning, can be complementary to traditional methods for electromagnetic field (EMF) exposureassessment, overcoming limitations and allowing extraction of new/deeper information. In this study, two examples of innovative EMF exposure assessment approaches are presented: (i) a stochastic approach based on low rank tensor approximations to assess indoor exposure to WLAN access point with unknown location and (ii) an application of Machine Learning to characterize indoor residential exposures to ELF magnetic field in children by considering the type of electric networks near the child home, the age and type of the child home, the type of heating and the family size.