Marco Castangia

Marco Castangia is a Research Fellow at the Interuniversity Department of Regional and Urban Studies and Planning (DIST) at Politecnico di Torino. He has been a Course Collaborator for the Programming Techniques course in Computer Science Engineering since the 2020/21 academic year, contributing to teaching and curriculum development.

• Current Position: Research Fellow at DIST
• Teaching Role: Programming Techniques, Computer Science Engineering
• Email: marco.castangia@polito.it

Research Interests
Marco's work focuses on applying machine learning and deep learning techniques to energy systems and environmental challenges. His research spans non-intrusive load monitoring for energy disaggregation, anomaly detection in household appliances using unsupervised methods, and neural network-based solar radiation forecasting via satellite image analysis. Recent projects include interpretable flood prediction models using transformer architectures and operational mode clustering for appliances.

Scientific Contributions
His publications demonstrate expertise in integrating AI with energy informatics and environmental modeling. Notable works include unsupervised learning for load monitoring (2024), transformer networks for flood forecasting (2023), and variational autoencoders for appliance anomaly detection (2022).

• Awards: National and international patent for Method for anomaly detection in electric appliances

Collaborations
Marco collaborates with researchers like Enrico Macii (doctoral advisor), Edoardo Patti, Massimo Poncino, and international colleagues such as Awet Abraha Girmay and Christian Camarda.