Guest Speaker: Dan Assouline, Data Scientist at Unisanté

When: Thursday, December 9, 11am-noon ET (Zoom)

Register >

Summary:  A promising path forward to achieve sustainable development goals is an increased reliance on renewable sources of energy. Optimized use of these energy sources, however, requires assessment of their potential supply, along with demand loads at locations of interest. This talk presents two projects that aim to use Machine Learning methods for renewable energy modeling. The first develops a general data-driven strategy that combines geographic information systems and machine learning methods to map large-scale energy potential for multiple renewable energy sources, particularly solar PV, with an application to Switzerland. The second aims to model electricity demand curves using time series analysis and machine learning models for clustering and anomaly detection, with an application to Swiss commercial building data. The general methodologies and results will be presented after a quick introduction to the machine learning concepts and models used in both projects.