A0849
Title: Explaining SDGs: How predict energy poverty in Italian households
Authors: Cecilia Camporeale - ENEA (Italy)
Giuseppe Scandurra - Parthenope University of Naples (Italy) [presenting]
Abstract: Energy poverty is a multifaceted and pervasive issue that poses significant challenges to policymakers and affected households. EP, understood as the inability of households to secure sufficient energy services for maintaining an adequate standard of living, makes it difficult to reconcile the objectives of the so-called energy trilemma, the simultaneous pursuit of energy security, environmental sustainability, and socially equitable access to energy sources. As climate change impacts intensify, addressing energy poverty becomes crucial for building resilient communities and fostering a sustainable future. EP measurement is generally based on different class indicators, which can be distinguished from households' perceptions of energy poverty and expenditure-based approaches. A classification algorithm is proposed that identifies Italian families living in energy deprivation using the data from the Household Budget Survey led by the Italian National Institute of Statistics (ISTAT). While exploring various machine learning algorithms, most analysis uses a random forest classifier. The proposed model can accurately classify families in energy poverty, providing valuable insights into the most important variables in predicting the risk of experiencing energy-related distress. In this way, it is possible to identify the levers on which policymakers should act to tackle energy poverty effectively.
