English Russian


Machine Learning for Automated Seismic Facies Classification in Paradigm 17 - Perth

February 07, 2017
Paradigm Office
Level 1
225 St. Georges Terrace
Perth, Australia


Presented by: Kamal Hami-Eddine

Breakfast Talk - Tuesday, February 7

7:30 AM to 8:00 AM - Networking Breakfast
8:00 AM to 9:00 AM - Presentation


Attendance is complimentary, and breakfast will be served.


Machine Learning for Automated Seismic Facies Classification in Paradigm 17
The further evolution of Stratimagic® and SeisFacies®

Machine learning is revolutionizing our lives, in areas as disparate as self-driving cars to facial recognition algorithms that now score higher recognition scores than live human viewers.  Seismic interpretation is also being altered by machine learning.  Below is a 3D lithofacies volume which was created by machine learning applied to a 3D prestack seismic volume, using the Rock Type Classification algorithm. Previously only a service offering, this is now a new software offering in Paradigm17.

Rock Type Classification calibrated to wells on a carbonate reef reservoir.

Within a few minutes, Rock Type Classification had classified, using facies logs at well locations, a 3D prestack seismic volume, and had returned a 3D volume consisting of rock classes: silicoclastics, bioherm (tight, wet or oil), limestone, shaly limestone, interbedded, and biostrom (tight, wet or oil).  It enabled the user to sidestep a traditional QSI (AVO & Inversion) project, which would have taken longer to complete.  Rock Type Classification is very easy to use, and because it is fully integrated with the 3D Canvas interpretation workspace, the user doesn’t have to open another application.

Waveform Classification (part of the Stratimagic product) has also been fully integrated with the 3D Canvas interpretation workspace in Paradigm 17, and it is now possible to do the classification task while engaged in interpretation, without interruption.

We invite you to join the presentation on how machine learning is changing the seismic interpretation landscape, and see demonstrations of the new Rock Type Classification and Waveform Classification algorithms, integral parts of the 3D Canvas workspace in Paradigm 17.


Kamal-Hami-Eddine.jpgKamal Hami-Eddine is Product Manager at Paradigm, and one of the inventors of Rock Type Classification. He served as research scientist and technical consultant for the company prior to assuming the current role. He is passionate about getting information out of data, and focusing on finding new solutions to existing customer challenges. He is an advocate of using statistics, analytics and machine learning in G&G workflows. He believes that big data and recent evolutions in machine learning, will soon enable O&G industry decision makers with an unprecedented level of valuable business information. He holds a master degree in Applied Mathematics and Stochastic Computations from the “Institut National des Sciences Appliquées” of Toulouse in France. In his free time, Kamal loves to cook for his friends, travel with his friends and play basketball.