Watch videos of Presentations from SEG 2017
Using Attributes to Interpret the Environment of Deposition - A Video Course. Taught by Kurt Marfurt, Rocky Roden, and ChingWen Chen
Dr. Kurt Marfurt and Dr. Tom Smith featured in the July edition of AOGR on Machine Learning and Multi-Attribute Analysis
Rocky Roden and Ching Wen Chen in May edition of First Break - Interpretation of DHI Characteristics using Machine Learning
Seismic interpretation and machine learning by Rocky Roden and Deborah Sacrey, GeoExPro, December 2016

LOGIN / MY PROFILE | LOG OUT

Paradise training of SOM, PCA, multi-attribute analysis

SEG 2015 | Blaine Taylor

An excerpt:

Welcome to Paradise – First Steps (Preview). The three basic tools which are used in Paradise are: 1. Principal Component Analysis, otherwise known as PCA. 2. Self Organizing Maps, otherwise known as SOM’s. 3. The interactive 2 Dimensional Colormap, which allows you to examine the classification groups (Neurons in a Neural Network), for common patterns within very complex data sets, of many simultaneous attributes.

...read more >

Full Article Text...

Multi-Attribute Analysis

Slide 2

New Training Facility

Our new Training facility was opened on 9/30/2015 on the occasion of our open house for the new offices.

8584 Katy Freeway

Suite 400

Houston, Texas, 77024

Slide 3

Geoinsights Training Facility

We have a new training facility in our offices!

  • 8 user workstations with state-of-the-art computers and twin 27 inch High Definition monitors
  • We use two large monitors instead of the old fashioned projectors
  • Ergonomic seating conditions with plenty of room enhance the training experience

Slide 4

Training Overview - Existing Courses

“Paradise Self Paced Basic Training; Version 2.0”

a)Complete Basic Software usage Tutorial for Version 2.0 Paradise

b)Complete with Multi-Attribute Seismic data, Well data and Horizon data

c)Included with Software Evaluation and Software Purchase or Lease

d)Course requires 10 to 12 hours to complete on clients own workstation

Slide 5

Training Overview - Existing Courses

“Welcome to Paradise –First Steps; Version 2.0”

a)Two Day course at our training facility

b)Basic Software usage for Version 2.0 Paradise

c)Basic training regarding Multi-Attribute Seismic

d)Creation of PCA to ascertain Attribute “Recipes”

e)Creation of SOM to allow interactive classification recognition

Slide 6

Training Overview - First Quarter 2016 Courses

“Paradise Self Paced Basic Training; Version 3.0”

a)Complete Basic Tutorial for Version 3.0 Paradise

b)Complete with Multi-Attribute Seismic data, Well data and Horizon data

c)Included with Software Evaluation and Software Purchase or Lease

d)Course will require about 12 hours to complete on clients own workstation

Slide 7

Training Overview - First Quarter 2016 Courses Cont. 

“Welcome to Paradise –First Steps; Version 3.0”

a)Two Day course at our training facility

b)Basic Software usage with the Version 3.0 Paradise user interface and features

c)Basic training regarding Multi-Attribute Seismic

d)Creation of PCA to ascertain Attribute “Recipes”

e)Creation of SOM to allow interactive classification recognition

Slide 8

Training Overview - First Quarter 2016 Courses Cont.

“Welcome to Paradise –Learning to Run; Version 3.0”

a)Advanced level two day course with pre-requisite of Basic course or completed tutorial

b)Covers Additional Version 3.0 features

c)Advanced level PCA for understanding Geology, and analyzing attribute contributions

d)Advanced level SOM for classification of Multi-Attribute Seismic with well calibration using horizon bounding techniques

e)Advanced ColorMapmanipulation for import/export

Slide 9

Training Overview - First Quarter 2016 Courses Cont. 

“Welcome to Paradise –Basic Architecture for installation and Maintenance”

a)Half day course at our Training Facility

b)Covers Hardware requirements versus project data volumes

c)Covers Installation and update procedures

d)Covers Administrator and User setup

Slide 10

Training Prices in the New Training Center

Self Paced Tutorial is included with Paradise software

Standard price for “Welcome to Paradise –First Steps Version 2.0 ” is $1500.00

Introductory Price is $500.00

Slide 11

Course Syllabus –“Welcome to Paradise –First Steps” 

  1. Multiple Seismic Attributes 
    1. Early Work
    2. Rock Solid Attributes and Other Sources
    3. Golden Dataset Overview
  2. Import 2D Data for Line "3" to Paradise
  3. Import 3D Attributes for Golden into Paradise
  4. Import Wells into Paradise
  5. Import Horizons to Paradise
  6. Paradise User Settings
    1. User Administration 
    2. Neural Processor Configuration 
    3. Settings Tab
  7. What Paradise Does
  8. Classification by Principal Component Analysis
    1. Create Portfolio
    2. Create Minimal PCA (Principal Component Analysis)
    3. View PCA
  9. Classification by Self-Organizing Map
    1. Attribute Choice
    2. Create Minimal SOM (Self Organizing Map)
  10. Interactive Neuron Identification with 2D Colormap

Slide 12

Course Syllabus – “Welcome to Paradise – First Steps” 

  1. Create Many Attribute PCA
  2. View PCA Results
  3.  Create SOM (Self Organizing Map)
  4. View and Analyze SOM Results

Slide 13

Welcome to Paradise –First Steps (Preview) 

The three basic tools which are used in Paradise are:

  1. Principal Component Analysis, otherwise known as PCA
  2. Self Organizing Maps, otherwise known as SOM’s
  3. The interactive 2 Dimensional Colormap, which allows you to examine the classification groups (Neurons in a Neural Network), for common patterns within very complex data sets, of many simultaneous attributes.

Slide 14 - Welcome to Paradise – First Steps (PCA Tool) 

Slide 15 - Welcome to Paradise –First Steps (SOM and 2d ColorMap)

Slide 16

Course Preview – Attribute Discussion

3D Isometric view of Reflected Seismic Wave 

Add the Part of the Reflected Seismic which we can measure 

Slide 17

Seismic Attributes

3D Isometric view of Reflected Seismic Wave 

Add the Part of the Reflected Seismic which we can’t measure

Slide 18

Seismic Attributes

3D Isometric view of Reflected Seismic Wave 

And finally the total energy of the entire reflected waveform

Slide 19

Seismic Attributes

Virtually all the Seismic Attributes are created from these three “Basic” portions of the Reflected Seismic Wave, in either Post Stack or Pre Stack processing.

  • Instantaneous Attributes
  • Geometrical Attributes
    • Curvature
    • Coherency
  • AVO Attributes
  • Spectral Decomposition Attributes
  • Inversion Attributes

Slide 20 - Course Preview -Classification Using PCA

Slide 21

View SOM Results

  1. Slide the Probability Filter Slider to see its’ effect.
  2. Turn the Set Value to “Outside” and repeat sliding the probability Filter up and down.
  3. Note the presence of the Fluvial Channel making itself evident as you move Probability filter values up and down. Leave the Value set to Outside and the Filter set to a high value

Slide 22

View SOM Results

  1. With the cursor set to the pointer, (click on the tools, or toggle the spacebar), run the cursor over the ColorMapand notice the effects as we select a single Neuron or Classification bin. You will soon note that it takes several Neurons to describe the entire Channel complex.
  2. Choosing the “Invert” button on the ColorMapshows only the additional neurons as you choose them.