Interpretation, Attributes and Machine Learning
Next Course Date: September 18, 2018
Instructor(s): Sarah Stanley & Carrie Laudon
Location: Denver, Colorado
This introductory two day course to Paradise enables geoscientists and engineers to use the major workflows and become acquainted with the basic capabilities of the multi-attribute analysis platform. Students will be trained in applying Principal Component Analysis (PCA) to a group of attributes to identify the most significant attributes in the set, then running, viewing and analyzing Self-Organizing Maps (SOMs) on selected attributes using different neural network configurations.
Next Course Date: TBD
Instructor(s): Deborah Sacrey, Auburn Energy
The Interpretation Workshop aims to equip geoscientists to interpret in Paradise using multi-attribute analysis by focusing on at least two projects in different geologic settings. The course will highlight the importance of using multiple attributes to extract more information from the seismic response.
Course objectives and topics to be covered:
- Demonstrate the capabilities of Paradise vs. traditional interpretation tools.
- Instruct in the applications of Paradise in different geologic settings.
- Learn the details of interpreting in Paradise
- Work in 2D and 1D Colorbars
- Export to an interpretation package - Kingdom and Petrel
- Export a neural masks
Generation and Application of Seismic Attributes in Paradise
Next Course Date: TBD
Instructor(s): Blaine Taylor & Gary Jones
As the name indicates, this course exercises the Attribute Generator in Paradise to generate, understand, and apply the various classes of attributes available in the comprehensive Paradise attribute library. Those interested in taking this course should have taken the Paradise Essentials course, logged extensive experience with Paradise, and/or viewed the set of online Paradise training videos.
Day 1: Basic Attribute Review
- Instantaneous Attribute (19) Description, Interpretation Usage, Calculation equations, Generation and preliminary viewing of single attributes.
- Geometric Attribute (55) Description, Interpretation Usage, Calculation equations, Generation and preliminary viewing of single attributes.
- Spectral Decomposition Attribute (27) Description, Interpretation Usage, Calculation equations, Generation and preliminary viewing of single attributes.
- By the end of the Day 1, the above attributes have either been generated or will be completed overnight, and ready at the beginning of the second day for multi-attribute analysis.
The second day will instruct in simultaneous analysis of multiple attributes using Neural Network, Machine Learning technology with the aim of identifying geologic patterns in the Great Southern Basin 3D dataset from New Zealand. Consequently, the student must be familiar with how to use Paradise as a prerequisite to this course.
The capstone of the second day is selecting the optimum combination (set) of attributes for specific geological features.