My fellow students know this and I hope recruiters will never read this: I was never good at math in university. It was only later when it came to the application in actual geophysical problems that tensors, linear algebra, and differential equations clicked. Personally, I don’t recommend this, as it makes life unnecessarily hard.
Machine Learning is a tool that will become more prevalent in our everyday work within the next decade. We can use it as a tool like I did in the SEG ML challenge. There I started using scikit-learn a Python library for machine learning. This library makes machine learning so easy you can build your first model in three lines of code:
clf = GradientBoostingClassifier()
First, you define your classifier. Second, you feed it some labeled data. Third, you predict on unseen data.
The code above will give you “some” result. Good models need some fine tuning. Luckily the Python community is all about using the code. Therefore, at the scientific Python conference SciPy a one-day tutorial was held. You can follow the tutorial as if you were there. The morning session is available here:
The afternoon session is available here:
and all the material to the workshop can be followed along and tweaked on GitHub.
Getting the math right
I love Scikit-Learn and it got me into machine learning with an extremely low barrier. However, deep learning is mostly about neural networks. These are a little more involved and doing research with them may require some math. Wait! Don’t go yet. If you never got Linear Algebra in school or uni, definitely keep reading. If you did, just wait for the next post that will delve deeper into deep learning for geoscientists.
On my journey through all the fascinating free courses, I will show you later, I was pointed toward 3Blue1Brown
He makes amazing math videos that make you understand math on an intuitive level. Their “Essence of Linear Algebra” playlist made me truly understand matrix multiplication for the first time, although I have been using it for almost 10 years.
This will get you set up for future math that will make all the fun in machine learning and deep learning in geoscience accessible to you.
Definitely, subscribe to stay updated with the next posts on all the resources I found on my journey into ML.
Latest posts by Jesper Dramsch (see all)
- Research Talk — Deep Learning for 4D Pressure Saturation Inversion [Youtube] - 2019-04-18
- Geysers in Slow Motion - 2019-02-04
- Keynote Bonanza and No Coffee – The EAGE / PESGB ML Workshop - 2018-12-17