Annual Winter School "Learning From Data" in Norway, 20th January 2019
 

During 20.01.2019 - 25.01.2019, we had the opportunity to attend the annual winter school held in Geilo, Norway. The theme of the winter school was 'Learning from data'. With the advent of sensors that frequently capture a snapshot of the physical world around us, a lot of data is being generated. Often the data captured is 'noisy'. As a result, new challenges arise on how to interpret the enormous amount of data collected. How can existing mathematical models of physical systems benefit from data? When the model predictions and actual data collected do not agree with each other what should be done? How can data be assimilated into models and improve them? Is the ‘value of information’ available in data worth the cost of collecting data?

In order to help us understand these issues better, experts in their fields gave lectures on the following topics:
• Data assimilation
• Inverse modeling and parameter estimation
• Uncertainty quantification
• Value of information
• Machine learning

The poster session provided us with the opportunity to interact with other researchers and learn more about the current research trends in various interdisciplinary fields. The winter school also included a data challenge where participants could apply machine learning algorithms to predict the intensity of a tropical storm and provide a 24-hour forecast based on the previous data collected.

Authors: Vijayasarathi Janardhanam, Alejandra Jayme and Sotirios Nikas