This video belongs to the openHPI course clean-IT: Towards Sustainable Digital Technologies. Do you want to see more?
To enable the transcript, please select a language in the video player settings menu.
About this video
Earth temperature measurements require the use of several stations across continents. These stations have high operational costs and computation requirement. We developed an AI that identifies subsets of stations suitable for future measurements, under budget constraints. As a result the algorithm identifies a fraction of the necessary sensor stations to calculation close to perfect measurement result, and saving a lot of computation at the same time. More information...
Francesco Quinzan is a Ph.D. candiate at the Hasso Plattner Institute in Potsdam, Germany. He works with the Algorithm Engineering Group, under the supervision of Prof. Dr. Tobias Friedrich. Previously, he studied pure mathematics at the University of Roma Tre in Rome, where he graduated with honours. His research focuses on Big Data.
Every day, an astonishing amount of online information is produced and collected in data-sets that are massive in size. Handling large data-sets requires automated systems that detect pattern in data, predict future outcome, and perform decision making under uncertainty. These systems rely on computational methods to perform feature selection, regression and classification.