Machine Learning in Geomechanics

This course is meant as a brief introduction to machine learning as it applies to geomechanics. We will start with an introduction to continuum and distinct-element numerical modeling. Then machine learning and surrogate modeling will be introduced and explained. We will discuss the advantages of each methodology and highlight the capability of numerical simulations to empower machine learning. Then we will present two real-world projects that involved the use of machine learning.

Course curriculum

    1. Introduction

    1. Introduction

    2. Why Discuss Numerical Modeling?

    3. What is Numerical Modeling?

    4. Numerical Modeling Methods & Software

    5. Explicit and Implicit Methods

    6. Continuum Modeling Advantages and Limitations

    7. Discontinuum Modeling Advantages and Limitations

    8. When To Use Numerical Models

    9. Model Simplification

    10. Model Size and Boundaries

    11. Workflow for Numerical Analysis

    12. Final Thoughts

    1. Introduction

    2. What is Machine Learning?

    3. Machine Learning Framework

    4. Feature Encoding & Engineering

    5. Model Selection & Training

    6. Model Evaluation & Tuning

    7. Utilizing Numerical Models & Machine Learning

    8. Data Generation & Sampling

    9. How to Get Started

    10. Final Thoughts

    1. Introduction

    2. Background

    3. Project Motivation

    4. Data & Preprocessing

    5. Machine Learning Model

    6. Analysis

    7. Conclusions

    1. Introduction

    2. Background

    3. Why Build a Surrogate Model?

    4. Numerical Model Development

    5. Design of Experiments

    6. Training A Surrogate Model

    7. Informing the Discontinuum Model

    8. Conclusion

    1. Feedback

About this course

  • Free
  • 39 lessons
  • 2 hours of video content

Instructors

Chris Thielsen

Geomechanics Engineer

Chris is a geomechanics engineer with experience in the numerical modeling of soil and rock and specializes in the development of machine learning models in the field of geomechanics. His current work focuses on the creation of surrogate models with artificial neural networks and training random forest models on core logging data that are used to predict rock strength at open-pit and underground mines.

Wei Fu

Geomechanics Engineer

Dr. Fu’s experience centers on computational, analytical, and experimental geomechanics in subsurface resource development (petroleum, geothermal energy, and mining) and underground excavation. He is experienced in the computational modeling of coupled geomechanics and fluid flow. He has experience in developing analytical models using fracture mechanics, designing and performing laboratory hydraulic fracturing and rock mechanics experiments, characterizing subsurface features, and integrating large field datasets with numerical simulations. Dr. Fu has worked on a variety of geomechanics projects, such as numerical modeling and analytical upscaling of hydraulic fracture swarms, experimental study of time-dependent fracture initiation, impact of natural fracture heterogeneities on hydraulic fracture propagation integrating discrete element modeling, laboratory experimentation, fracture mechanics modeling, and numerical analyses of dynamic tunneling processes and tunnel durability.

Social proof: reviews

5 star rating

Very nice introductory course

Mussie Kidane

The course really gives you a good introduction to machine learning in geomechanics.

The course really gives you a good introduction to machine learning in geomechanics.

Read Less
5 star rating

Excellent Short Course!

Jacob Sibanda

You guys rock! And thank you for the material you shared (on the course and on github), thank you for sharing your knowledge as well! :)

You guys rock! And thank you for the material you shared (on the course and on github), thank you for sharing your knowledge as well! :)

Read Less
5 star rating

Great

Jesse Nyokabi

Excellent course

Excellent course

Read Less