Datenwissenschaft in der Erdbeobachtung

What it is about

You deepen your knowledge in data science for Earth observation and learn central methods of machine learning (e.g., SVM, Random Forests, Deep Neural Networks, CNNs, RNNs). You apply these methods to typical EO tasks such as classification, segmentation and regression and finally develop a concrete ML model as a project work.

What you will be able to do

• Fundamental understanding of typical ML methods (e.g., SVM, Random Forest, Deep Neural Networks)
• Understanding the applications of these ML methods in Earth observation
• Distinction between physical models and data-driven ML models as well as their advantages/disadvantages
• Knowledge of methods for quantifying uncertainties in Deep Neural Networks
• Understanding a typical ML workflow (labeling, architecture design, training, fine-tuning, validation)
• Selection and implementation of appropriate models for specific EO problems; design of experiments and validation
• Production of precise scientific research reports

What the module consists of

• VorlesungVermittlung der theoretischen Grundlagen und Präsentationen
• ÜbungenVertiefung durch praktische Aufgaben, Partner-/Gruppenarbeit und Problemlösung
• HackathonKonsolidierung des theoretischen Wissens anhand typischer EO‑Problemlösungen
• ProjektarbeitEntwicklung eines konkreten ML‑Modells in Teams (2–3 Personen)

Teaching method

• Vortrag/PräsentationEinführung und Erklärung der Inhalte
• Fragend‑entwickelnde MethodenInteraktion mit Studierenden anregen
• Übungsbeispiele in Partner-/GruppenarbeitAnwendung theoretischer Grundlagen auf praktische Aufgaben
• Hackathon (praktisch, Python)Praxisnahe Vertiefung typischer Problemlösungen in der Erdbeobachtung