Quick Test

Introduction to Machine Learning for Geoscientists

Best Seller

By Sami Saad ElKurdy

The Course Instructor

Sami Saad ElKurdy
Exploration/Development Geophysicist & Machine Learning Consultant
  • Instructor Rating : N/A
  • Reviews : 0
  • Students : 28
  • Courses : 1

▪ Over 40 years of exploration and development experience in the Middle East, North and East Africa, Europe, New Zealand, South America, South East Asia and Canada.

▪ Have a good track record of play concept development and prospect/ lead generation within a teamwork environment.

▪ Structural seismic interpretation in rift basins, overthrust belts wrench controlled and salt structure areas using both 2D and 3D seismic.

▪ Stratigraphic seismic interpretation using both sequence stratigraphy and trace attributes identifying both clastics and carbonates anomalies.

▪ Experienced in various interpretation workstations, e.g. Landmark (Decision Space Desktop, Well Seismic Fusion &GeoProbe), Paradigm, IESX, SeisX, and Winpics. Also experienced in various mapping systems, e.g. Zmap+, Petrosys, GES, and Surfer.

▪ Generating structural and stratigraphic maps, isopach and time thickness maps, velocity maps. Depth conversion using geostatistics and mapping systems.

▪ Tying synthetic seismograms and VSPs to seismic.

▪ Time Lapse (4D) seismic interpretation and analysis.

▪ Performed AVO studies and pre stack depth migration. Seismic Inversion using Hampson-Russell Strata.

▪ Seismic attributes extraction and neural network mapping..

What you'll learn

  1. Learn the different kinds of Machine Learning: Supervised, Unsupervised and Reinforcement Learning
  2. Learn the various packages to apply ML, e.g. Python, pandas, numpy, matplotlib, etc...
  3. Learn data wrangling and exploratory data analyses (EDA)
  4. Learn industry standard practices for workflows to build ML models, e.g. train-validate-test split
  5. Learn how to build simple to gradient boosting to neural network models.
  6. Disciplines that will benefit from the course are Geologists, Geophysicists, petrophysicists, and static reservoir modellers.
  7. Both seismic and well data will be provided by the instructor from open sources.
  8. The course will be structured as a practical workshop, where the attendees will be able to follow along by applying the same cells in the jupyter lab notebooks supplied.
  9. The course will cover a crash course in python programming language.
  10. A sample segy 3D will be worked on to develop an unsupervised model. This might be of interest to geophysicists.
  11. A sample set of wells will be worked on to develop a supervised regression/classification model. This might be of interest to petrophysicists and geologists and reservoir modellers.


You need to install several open source packages as descriped below and illustrated in the recorded video:

  • Python ecosystem: will try to install winpython 10.3 version of python. This will automatically install the main packages, including jupyter lab.
  • Will install Opendtect. This will only be available on computers with 16GB of RAM. If less, we cannot install.
  • Will install ydata-profiling for exploratory data analysis.
  • Will test the system during the meeting to make sure that all is running.
  • Will install catboost
  • Install Seisee from DMNG which is a seismic viewer.
  • Install lasio for las logs processing.
  • Install segyio for seismic segy processing.

Similar Courses