Report from the first Training School

Report from the first Training School

The 1st BigSkyEarth Training School was held in the Institute for Remote Sensing Methodology at the German Aerospace Center DLR Oberpfaffenhofen in Germany.

The Training School gathered 33 trainees from 14 countries with expert instructors from Europe and the US for hands-on training on Big Data analytics.

Since the identified challenges are similar in astronomy and Earth Observation, with signal processing, statistics, machine learning, and computer science as the common denominator, the Training School has aimed at boosting the communication within and between disciplines and applications areas, by propagating and advancing relevant common solutions developed within Big Data analysis and management research and industrial environments.

The goal has been to share valuable know-how encompassing from sensor and data modelling, features extraction and metadata, information representation, data structures, pattern recognition, statistical/machine learning, data analytics, advanced visualisation, to data mining and KDD. Specific computer science topics have been addressed as particular programming techniques, cloud computing methodologies and related topics. The key aspect has been to address all these topics in synergy, setting them up in a logical interdisciplinary framework bridging the diverse areas addressed by the school.

The participants were selected for the school based on the quality and relevance of the research project descriptions provided as part of their application procedures. Their backgrounds ranged from astronomy — around 40% — to remote sensing — 55% — and computer science. The institutions they work for go from observatories to startup companies. About 20% were women, again from both astronomical and Earth observation backgrounds.

After acceptance in the school, they prepared for the attendance by going through the JPL-Caltech Summer School on the Coursera platform. After that they gathered at the Oberpfaffenhofen site of the German Aerospace Center DLR for a week of theoretical and practical work. Lessons were given on topics ranging from feature extraction for astronomical time series to large scale machine learning and visualization.

Speakers included Ashish Mahabbal from Caltech (USA), Mihai Datcu from the German Aerospace Center DLR (Germany), Brian Mac Namee from the School of Computer Science at University College Dublin (Ireland), Stefano Cavuoti from the Osservatorio Astronomico di Capodimonte (Italy), Marco Quartulli from Vicomtech-IK4 (Spain), Maria Gritsevich from the Finnish Geospatial Research Institute (Finland), Giovanni Nico from CNR IAC (Italy), Laurent Eyer from the University of Geneva (Switzerland) and Eric Feigelson from Penn State University (USA).

Cross-domain working groups of participants were established carrying forward specific data analysis projects focusing on regression tasks, high-dimensionality analytics, feature engineering and clustering, supervised classification and thematic mapping, and visualization. The material for the course — presentations and hands-on work notebooks by the speakers and the material being produced by the participant groups — is in the process of being published online.

TrainingSchool-photoList of trainees:

Name country
Gor Mikayelyan Armenia
Mihran Vardanyan Armenia
Praveen K. Pandey Belgium
Darko Zaric Croatia
Matti Aleksanteri Eskelinen Finland
Billy Braithwaite Finland
Zhang Chao France
Nestor Koueya France
Sina Montazeri Germany
Fathalrahman Adam Germany
Hu Jingliang Germany
Kersten Clauss Germany
Homa Ansari Germany
Margherita Grossi Germany
Patrick Hannawald Germany
Danfeng Hong Germany
Dimitris Marmanis Germany
Mohammad Omidalizarandi Germany
Melinda Nagy Hungary
Patricia O’Byrne Ireland
Luigi Mascolo Italy
Rafał Wojaczyński Poland
Justyna Średzińska Poland
Andreea Griparis Romania
Alexandru-Cosmin Grivei Romania
Aleš Marsetič Slovenia
Klemen Čotar Slovenia
Iñaki Arruabarrena Spain
Daniel Garabato Míguez Spain
Carlos de Hoyos Fernández de Córdova Spain
Maria Manolopoulou UK
Christopher Frohmaier UK