Introduction to essential ML libraries, data visulization tools & cloud services.
Course contents:
- Introduction to machine learning techniques
- Application of machine learning on the example of modern geophysical problems
- Supervised learning
- Unsupervised learning
- Deep neural networks
- best practices
Machine learning (ML) has become the main driver of numerous innovative applications: from automated driving and medical devices to industrial automation and electronics. The course offers a comprehensive introduction to machine learning techniques and accompanying examples of the application of machine learning to modern problems in the fields of petroleum engineering, microseismic monitoring, underground storage, and geothermal energy. The goal of the course is to provide trainees with a fundamental understanding of machine learning algorithms sufficient to apply them to solve problems. Topics cover classical supervised learning (linear regression, logistic regression, support vector machine), ensemble methods, neural networks (DNN, CNN, RNN, GAN), and unsupervised learning (clustering, principal component analysis, anomaly detection). Although machine learning typically requires HPC resources and advanced programming skills, the course is designed in a way that trainees only need basic programming skills in Python.
Ivan Abakumov received his Bachelor's (2009) and Master's (2012) degrees in Physics from St. Petersburg State University. While studying, Ivan was working as a teacher of computer science and led advanced math classes for school graduates. After receiving a Master's degree, Ivan completed internships at Microseismik s.r.o.. (Czech Republic) and Shell (The Netherlands). From 2012 till 2015, Ivan participated in a Shell-funded research project aimed at inverting velocity anomalies from time-lapse seismic data. Since late 2013, Ivan worked on a Ph.D. thesis at the University of Hamburg. His thesis covers the aspects of multidimensional stacking for seismic data quality enhancement. After earning his Ph.D. Ivan has worked on short-term research projects at the Trofimuk Institute of Petroleum Geology and Geophysics (Russia) and the Norwegian University of Science and Technology (Norway). Since September 2017, Ivan has served as a research scientist at Freie Universität Berlin where he developed innovative approaches for microseismic data inversion. At that time Ivan was already highly interested in machine learning techniques and open-source ML software. Since 2019, Ivan offers courses on machine learning and deep learning for master students. Since 2020, Ivan leads a Data Science team in one of the tech companies in Berlin. Ivan actively reviews manuscripts for Geophysics and EAGE on the topics of machine learning and digitalization.
This course is supported by:
Coffee, tea, water and lunch are included in the course fee for all course days.
Please bring your own computer.
Registration Deadline: January 31, 2023
Please register for your binding registration and you will receive an invoice for the course fee after registration. If you cancel your registration by February 15, 2023, DGMK will refund any fees already paid minus a processing fee of EUR 30.-. After February 15, 2023, the full course fee must be invoiced.
By public transportation
From Hauptbahnhof (Main Train Station) to the DGMK Office please use S1, S2,S3 to station 'Königsstraße'. Take the exit in direction 'Struenseestraße'. Follow the street 'Behnstraße' in direction to the river 'Elbe'.
By car
There are several car parks nearby.
Underground car park Holzhafen : Große Elbstraße 61/63, 22767 Hamburg
Parking lot Cruise Center : Van-der-Smissen-Straße 1-4, 5, 22767 Hamburg
Registration is closed.
Head of the Geo-Energy-Systems and Subsurface Technologies Department
Coordination of the Department Geo-Energy-Systems and Subsurface Technologies