This book provides readers with a timely review and discussion of the success, promise, and perils of machine learning in geosciences. It explores the fundamentals of data science and machine learning, and how their advances have disrupted the traditional workflows used in the industry and academia, including geology, geophysics, petrophysics, geomechanics, and geochemistry. It then presents the real-world applications and explains that, while this disruption has affected the top-level executives, geoscientists as well as field operators in the industry and academia, machine learning will ultimately benefit these users. The book is written by a practitioner of machine learning and statistics, keeping geoscientists in mind. It highlights the need to go beyond concepts covered in STAT 101 courses and embrace new computational tools to solve complex problems in geosciences. It also offers practitioners, researchers, and academics insights into how to identify, develop, deploy, and recommend fit-for-purpose machine learning models to solve real-world problems in subsurface geosciences.
Author(s): Shuvajit Bhattacharya
Series: SpringerBriefs in Petroleum Geoscience & Engineering
Publisher: Springer
Year: 2021
Language: English
Pages: 189
City: Singapore
Preface
Acknowledgments
Contents
About the Author
Acronyms
1 Introduction
1.1 What are Big Data, Data Analytics, and Machine Learning?
1.1.1 Big Data
1.1.2 Data Analytics
1.1.3 Machine Learning
1.2 History of Machine Learning
1.3 Where are the Geoscientists in this Digital Age and ML-Tsunami?
1.4 Why should we care about Machine Learning in Geosciences?
1.5 Types of Data Analytics
1.6 Geoscience Databases
1.6.1 Numerical Data Types
1.6.2 Non-Numerical Data Types
1.7 Scales, Resolutions, and Integration of Common Geologic Data
References
2 A Brief Review of Statistical Measures
2.1 Random Variable
2.2 Common Types of Geologic Data Analysis
2.2.1 Univariate Analysis
2.2.2 Bivariate Analysis
2.2.3 Time Series Analysis
2.2.4 Spatial Analysis
2.2.5 Multivariate Analysis
References
3 Basic Steps in Machine Learning-Based Modeling
3.1 Identification of the Problem
3.2 Learning Approaches
3.2.1 Unsupervised Learning
3.2.2 Supervised Learning
3.2.3 Semi-Supervised Learning
3.3 Data Pre-Processing
3.3.1 Data Integration and Feature Selection
3.3.2 Data Cleansing
3.3.3 Statistical Imputation for Missing Data
3.3.4 Data Abstraction
3.3.5 Feature Engineering
3.4 Data Labeling
3.5 Machine Learning-Based Modeling
3.5.1 Data Splitting
3.5.2 Model Training
3.5.3 Model Validation and Testing
3.6 Model Evaluation
3.6.1 Quantification of Model Performance and Error Analysis
3.6.2 Model Complexity
3.7 Model Explainability
3.7.1 Sensitivity Analysis or Key Performance Indicators (KPI)
3.7.2 Partial Dependence Plots
3.7.3 SHapley Additive exPlanations
3.7.4 Local Interpretable Model-Agnostic Explanations
3.8 Knowledge Discovery, Presentation, and Decision-Making
References
4 A Brief Review of Popular Machine Learning Algorithms in Geosciences
4.1 K-means Clustering
4.2 Artificial Neural Network
4.2.1 Hidden Layer
4.2.2 Learning Rate
4.2.3 Momentum
4.2.4 Activation Function
4.3 Support Vector Machine
4.4 Decision Tree and Random Forest
4.4.1 Decision Tree
4.4.2 Random Forest
4.5 Bayesian Network Theory
4.6 Convolutional Neural Network
4.6.1 Fully Connected Network
4.6.2 Encoder-Decoder Network
4.6.3 Optimizing CNNs
4.6.4 Strategies to consider in CNN Modeling
4.7 Recurrent Neural Network and Long Short-Term Memory
4.8 Ensemble Approach
4.9 Physics-Informed Machine Learning
References
5 Summarized Applications of Machine Learning in Subsurface Geosciences
5.1 Outlier Detection
5.2 Petrophysical Log Analysis
5.2.1 Facies Clustering and Classification
5.2.2 General Rationale behind the use of Conventional Well Logs for Facies Identification
5.2.3 Machine Learning for Well-Log-Based Facies Clustering and Classification
5.2.4 Fracture Classification
5.2.5 General Rationale behind the use of Conventional Well Logs for Fracture Classification
5.2.6 Machine Learning for Well-Log-Based Fracture Classification
5.2.7 Well-Log-Based Rock Property Prediction
5.3 Seismic Data Analysis
5.3.1 General Rationale behind the use of Seismic Attributes in ML Applications
5.3.2 Machine Learning for Seismic Facies Clustering and Classification
5.3.3 Fault Classification
5.3.4 Seismic-Based Rock Property Prediction
5.4 Fiber-Optic-Based Fluid Flow Prediction
5.5 Rock Characterization (Core, Outcrop, Petrography, and Geochemistry)
References
6 The Road Ahead
