My Path from Geoscience into Machine Learning

Transcript

1. A Path from Geoscience into ML Jesper Dramsch
2. If You Take One Thing Away Become a T-shaped person and stay curious. Become knowledgeable in many topics and an expert in a few. Make connections and share what you learn. Build a portfolio of projects and probably learn coding.
3. Where are we going today? • My Background • What’s Machine Learning? • Geo to ML?! • What’s important in ML?
4. My Path
5. My Path BSc in Geophysics Hamburg, Germany 2011
6. My Path MSc in Geophysics Hamburg, Germany 2014 BSc in Geophysics Hamburg, Germany 2011
7. My Path MSc in Geophysics Hamburg, Germany 2014 PhD in Geophysics Technical University of Denmark 2019/2020 BSc in Geophysics Hamburg, Germany 2011
8. My (linearized) Path MSc in Geophysics Hamburg, Germany 2014 PhD in Geophysics (ish) Technical University of Denmark 2019/2020 BSc in Geophysics Hamburg, Germany 2011
9. My Path MSc in Geophysics Hamburg, Germany 2014 PhD in Geophysics (ish) Technical University of Denmark 2019/2020 Machine Learning Engineer GMV 2020! BSc in Geophysics Hamburg, Germany 2011
10. My Path MSc in Geophysics Hamburg, Germany 2014 PhD in Geophysics (ish) Technical University of Denmark 2019/2020 Machine Learning Engineer GMV 2020! BSc in Geophysics Hamburg, Germany 2011
11. My Path MSc in Geophysics Hamburg, Germany 2014 PhD in Geophysics (ish) Technical University of Denmark 2019/2020 Machine Learning Engineer GMV 2020! BSc in Geophysics Hamburg, Germany 2011 The interesting bit.
12. Why switch careers? Rocks are nice you know.
14. BSc in Geophysics
15. MSc in Geophysics BSc in Geophysics
16. MSc in Geophysics BSc in Geophysics PhD in 2020
17. Don’t need that in my life (Also climate change...)
18. What is Machine Learning? And why should I care?
20. if x > 5 Classical rule-based and expert systems
21. Computers figure it out based on the data
22. Models that do not have external information Blackbox models
23. Some seismic data
24. An “inversion”
25. Self-driving cars
26. Neural Networks The new hype that is older than all of us
27. A neural network
28. A neural network
29. What is AI, ML, and Deep Learning?
30. A neural network as it is used today
31. The learned filters
32. What is the connection Why do we care in geoscience?!
33. Matching of seismic data
34. The differences between unmatched and matched
35. Compared to a baseline method
36. • Generating Synthetic Data • Earthquake localisation • Drill Core Processing • Clustering of Well Data • Prediction of Volcanic activity • Satellite Data What else though?
37. How did I change careers?
38. Start Projects Application > Theory • Participate in Hackathons • Build Apps • Compete on Kaggle • Get familiar with the Math
39. Build a breadth of knowledge Dive deep in some expertise