Application of Artificial Neutral Network and Geographic Information System to Evaluate Retrofit Potential in Public School Buildings

School buildings in Italy are old, in critical maintenance conditions and they often perform below acceptable service levels. Nevertheless, data to guide renovation policies are missing or very expensive to retrieve. This paper presents a methodology for evaluating building’s energy savings potential, using the Certificazione Energetica degli Edifici (CENED) database, concerning energy performance labelling. Data are first clustered to identify most common thermo-physical properties. Three retrofit scenarios are then defined and energy savings potential, for each of the three, is evaluated through eight neural networks. Ultimately, data are geocoded and further processed to guide the definition of the retrofit strategy in most critical areas in Lombardy region. The results of the three scenarios proved that the highest energy savings can be obtained through retrofit interventions on around 50% of buildings. In conclusion, further insights on retrofit costs analysis and future development of the research are discussed.