Future Frontiers: AI and Quantum Computing in Oil & Gas Exploration
COURSE OBJECTIVE
The oil and gas industry is looking to adopt machine learning technology to solve its problems and improve business outcomes. However, many of these companies struggle to do so because they lack the necessary knowledge to effectively use the technology. As a result, they are unable to obtain reliable and efficient results from machine learning models, which impacts their investments and overall performance.
The petroleum industry (oil and gas) is divided into upstream, midstream and downstream. Upstream summarizes the subsurface (mining) part of the industry, including exploration followed by field development and crude oil/gas production. Midstream represents the transportation of oil and gas, and downstream is the refining, that is, the production of fuels, lubricants, plastics and other products. Artificial intelligence AI solutions are already applied and their results.
During the course, participants learn to apply machine learning to seismic processing and interpretation problems. The course covers automated salt interpretation and briefly introduces strategies for solving other seismic interpretation problems.
The course explains some important aspects of the application of deep learning algorithms to seismic images. And the choice of image size and image preprocessing is explained.
The course discusses issues related to geological modeling, including supervised machine learning for estimating petrophysical properties away from well locations.
We will be discussing a module called geomodeling which is used to understand the structural framework of a reservoir. This is done by analyzing the information obtained via seismic interpretation. To estimate petrophysical property values on a threedimensional mesh, machine learning can be employed. Variograms serve as the foundation of almost all 3D spatial modeling techniques that are relevant to the industry. In this course, we will be using the Gaussian process regression model, which is a part of the probabilistic artificial intelligence, for the purpose of geomodeling.
The course aims to demonstrate various approaches for developing machine learning models to analyze decline curves and forecast oil production. It covers advanced machine learning forecasting models, including recurrent neural networks, long shortterm memory, Bayesian machine learning, genetic algorithms, and advanced ensemble learning models.
Production modeling using machine learning methodologies, including production optimization, is discussed. A methodology for virtual measurement and the formulation of virtual sensors is explained.
The course enables participants to leverage generative AI algorithms and quantum computing techniques to enhance exploration and reservoir characterization in the oil and gas upstream sector. By utilizing advanced technologies, participants can generate more accurate subsurface models and gain valuable insights into reservoir properties, leading to better decisionmaking and maximizing hydrocarbon recovery.
QUANTUM COMPUTING
Quantum Machine Learning is the integration of quantum algorithms within Machine Learning programs. Machine learning algorithms are used to calculate large amounts of data, quantum machine learning uses qubits and quantum operations or specialized quantum systems to improve the speed of calculation and data storage performed by a program's algorithms. For example, some mathematical and numerical techniques from quantum physics are applicable to classical deep learning. A quantum neural network has computational capabilities to decrease the number of steps, the qubits used, and the computing time.
The important objective of the course is to show the use of quantum computing and tensor networks to improve the calculation of machine learning algorithms.
Additionally, the course explains quantum computing, quantum circuits, important quantum algorithms, quantum mechanics, quantum error and correction, and quantum machine learning.
The course explains applications of quantum computing, such as Quantum Machine Learning, in oil production forecasting models. The improvement of quantum models over traditional ML models is presented, for example LSTM neural networks versus LSTM quantum neural networks.
IMPORTANT
In the oil and gas upstream industry, there is a pressing need to accurately apply both traditional and quantum artificial intelligence. To address this, we have developed an advanced validation module along with powerful model risk techniques and probabilistic machine learning methodologies. These tools help us to understand the uncertainties that exist in the results we obtain. Additionally, we have included a module called XAI which ensures that the models we use are not "black boxes" and can be interpreted.
WHO SHOULD ATTEND?
The course is designed for professionals in the oil and gas industry who want to enhance their skills in building effective artificial intelligence and quantum computing models for upstream applications. Participants are expected to have knowledge of statistics to better comprehend the topics covered in the course.
Schedules:

Europe: MonFri, CEST 1619 h

America: MonFri, CDT 1821 h

Asia: MonFri, IST 1821 h
Price: 6 900 €
Level: Advanced
Duration: 36 h
Material:

Presentations PDF

Exercises in Excel, R, Python, Jupyterlab y Tensorflow
AGENDA
AI and Quantum Computing in
Oil and Gas UpStream
Machine Learning
Module 1: Machine Learning

Definition of Machine Learning

Machine Learning Methodology

Data Storage

Abstraction

Generalization

Assessment


Supervised Learning

Unsupervised Learning

Reinforcement Learning

Deep learning

Typology of Machine Learning algorithms

Steps to Implement an Algorithm

Information collection

Exploratory Analysis

Model Training

Model Evaluation

Model improvements

Machine Learning in consumer credit risk


Machine Learning in credit scoring models

Quantum Machine Learning
Module 2: EDA Exploratory Analysis

Data typology

Transactional data

Unstructured data embedded in text documents

Social Media Data

Data sources

Data review

Target definition

Time horizon of the target variable

Sampling

Random Sampling

Stratified Sampling

Rebalanced Sampling


Exploratory Analysis:

Histograms

Q Q Plot

Moment analysis

boxplot


Treatment of Missing values

Multivariate Imputation Model


Advanced Outlier detection and treatment techniques

Univariate technique: winsorized and trimming

Multivariate Technique: Mahalanobis Distance


Exercise 1: EDA Exploratory Analysis
Module 3: Feature Engineering

Data Standardization

Variable categorization

Equal Interval Binning

Equal Frequency Binning

ChiSquare Test


Binary coding

Binning

Kind of transformation

Univariate Analysis with Target variable

Variable Selection

Treatment of Continuous Variables

Treatment of Categorical Variables

Gini

Information Value

Optimization of continuous variables

Optimization of categorical variables


Exercise 2: Detection and treatment of Advanced Outliers

Exercise 3: Stratified and Random Sampling in R

Exercise 4: Multivariate imputation model

Exercise 5: Univariate analysis in percentiles in R

Exercise 6: Continuous variable optimal univariate analysis in Excel

Exercise 7: Estimation of the KS, Gini and IV of each variable in Excel

Exercise 8: Feature Engineering of variables
Unsupervised Learning
Module 4: Unsupervised models

Hierarchical Clusters

K Means

Standard algorithm

Euclidean distance

Principal Component Analysis (PCA)

Advanced PCA Visualization

Eigenvectors and Eigenvalues

Exercise 9: Segmentation of the data with KMeans R
Supervised Learning
Module 5: Logistic Regression and LASSO Regression

Econometric Models

Logit regression

probit regression

Piecewise Regression

survival models


Machine Learning Models

Lasso Regression

Ridge Regression


Model Risk in Logistic Regression

Exercise 10: Lasso Logistic Regression in R

Exercise 11: Ridge Regression in R
Module 6: Trees, KNN and Naive Bayes

Decision Trees

Modeling

Advantages and disadvantages

Recursion and Partitioning Processes

Recursive partitioning tree

Pruning Decision tree

Conditional inference tree

Tree display

Measurement of decision tree prediction

CHAID model

Model C5.0


KNearest Neighbors KNN

Modeling

Advantages and disadvantages

Euclidean distance

Distance Manhattan

K value selection


Probabilistic Model: Naive Bayes

Naive bayes

Bayes' theorem

Laplace estimator

Classification with Naive Bayes

Advantages and disadvantages


Exercise 12: KNN and PCA
Module 7: Support Vector Machine SVM

Support Vector Classification

Support Vector Regression

Optimal hyperplane

Support Vectors

Add costs

Advantages and disadvantages

SVM visualization

Tuning SVM

Kernel trick

Exercise 14: Support Vector Machine in R
Module 8: Ensemble Learning

Classification and regression ensemble models

Bagging

Bagging trees

Random Forest

Boosting

Adaboost

Gradient Boosting Trees

Xgboost

Advantages and disadvantages

Exercise 15: Boosting in R

Exercise 16: Bagging in R

Exercise 17: Random Forest, R and Python

Exercise 18: Gradient Boosting Trees
Deep Learning
Module 9: Introduction to Deep Learning

Definition and concept of deep learning

Why now the use of deep learning?

Neural network architectures

Feedforward network

R deep learning

Python deep learning

Convolutional Neural Networks

Use of deep learning in image classification

cost function

Gradient descending optimization

Use of deep learning

How many hidden layers?

How many neurons, 100, 1000?

How many times and size of the batch size?

What is the best activation function?


Hardware, CPU, GPU and cloud environments

Advantages and disadvantages of deep learning
Module 10: Deep Learning Feed Forward Neural Networks

Single Layer Perceptron

Multiple Layer Perceptron

Neural network architectures

Activation function

Sigmoidal

Rectified linear unit (Relu)

The U

Selu

Hhyperbolic hypertangent

Softmax

Other


Back propagation

Directional derivatives

Gradients

Jacobians

Chain rule

Optimization and local and global minima


Exercise 19: Deep Learning Feed Forward
Module 11: Deep Learning Convolutional Neural Networks CNN

CNN for pictures

Design and architectures

Convolution operation

Descending gradient

Filters

Strider

Padding

Subsampling

Pooling

Fully connected

Exercise 20: deep learning CNN
Module 12: Deep Learning Recurrent Neural Networks RNN

Natural Language Processing

Natural Language Processing (NLP) text classification

Long Term Short Term Memory (LSTM)

Hopfield

Bidirectional associative memory

Descending gradient

Global optimization methods

Oneway and twoway models

Deep Bidirectional Transformers for Language Understanding

Exercise 21: Deep Learning LSTM
Module 14: Generative Adversarial Networks (GANs)

Generative Adversarial Networks (GANs)

Fundamental components of the GANs

GAN architectures

Bidirectional GAN

Training generative models

Exercise 22: Deep Learning GANs
Module 15: Tuning Hyperparameters

Hyperparameterization

Grid search

Random search

Bayesian Optimization

Train test split ratio

Learning rate in optimization algorithms (e.g. gradient descent)

Selection of optimization algorithm (e.g., gradient descent, stochastic gradient descent, or Adam optimizer)

Activation function selection in a (nn) layer neural network (e.g. Sigmoid, ReLU, Tanh)

Selection of loss, cost and custom function

Number of hidden layers in an NN

Number of activation units in each layer

The dropout rate in nn (dropout probability)

Number of iterations (epochs) in training a nn

Number of clusters in a clustering task

Kernel or filter size in convolutional layers

Pooling size

Batch size

Exercise 23: Optimization Xboosting, Random forest and SVM

Exercise 24: Optimized Deep Learning
Probabilistic Machine Learning
Module 16: Probabilistic Machine Learning

Introduction to probabilistic machine learning

Gaussian models

Bayesian Statistics

Bayesian logistic regression

Kernel family

Gaussian processes

Gaussian processes for regression


Hidden Markov Model

Markov chain Monte Carlo (MCMC)

Metropolis Hastings algorithm


Machine Learning Probabilistic Model

Bayesian Boosting

Bayesian Neural Networks

Exercise 25: Gaussian process for regression

Exercise 26: Bayesian Neural Networks
Model Validation
Module 17: Validation of traditional and Machine Learning models

Model validation

Validation of machine learning models

Regulatory validation of machine learning models in Europe

Out of Sample and Out of time validation

Checking pvalues in regressions

R squared, MSE, MAD

Waste diagnosis

Goodness of Fit Test

Multicollinearity

Binary case confusion matrix

KFold Cross Validation

Diagnostico del modelo

Exercise 27: Validación avanzada de la regression

Exercise 28: Diagnostico de la regresión

Exercise 29: KFold Cross Validation in R
Module 18: Advanced Validation of AI Models

Integration of stateoftheart methods in interpretable machine learning and model diagnosis.

Data Pipeline

Feature Selection

Blackbox Models

Posthoc Explainability

Global Explainability

Local Explainability

Model Interpretability

Diagnosis: Accuracy, WeakSpot, Overfit, Reliability, Robustness, Resilience, Fairness

Model comparison


Comparative for Regression and Classification

Fairness Comparison


Exercise 30: Validation and diagnosis of advanced credit scoring models
Auto Machine Learning and XAI
Module 19: Automation of ML

What is modeling automation?

That is automated

Automation of machine learning processes

Optimizers and Evaluators

Modeling Automation Workflow Components

Hyperparameter optimization

Global evaluation of modeling automation

Implementation of modeling automation in banking

Technological requirements

Available tools

Benefits and possible ROI estimation

Main Issues

Genetic algorithms

Exercise 31: Automation of the modeling, optimization and validation of pricing models
Explainable Artificial Intelligence
Module 20: Explainable Artificial Intelligence XAI

Interpretability problem

Machine learning models

1. The challenge of interpreting the results,

2. The challenge of ensuring that management functions adequately understand the models, and

3. The challenge of justifying the results to supervisors


Black Box Models vs. Transparent and Interpretable Algorithms

Interpretability tools

Shap, Shapley Additive explanations

Global Explanations

Dependency Plot

Decision Plot

Local Explanations Waterfall Plot


Lime, agnostic explanations of the local interpretable model

Explainer Dashboard

Other advanced tools

Exercise 32: XAI interpretability of pricing
Quantum Computing
Module 21: Quantum computing and algorithms
Objective: Quantum computing applies quantum mechanical phenomena. On a small scale, physical matter exhibits properties of both particles and waves, and quantum computing takes advantage of this behavior using specialized hardware. The basic unit of information in quantum computing is the qubit, similar to the bit in traditional digital electronics. Unlike a classical bit, a qubit can exist in a superposition of its two "basic" states, meaning that it is in both states simultaneously.

Future of quantum computing in insurance

Is it necessary to know quantum mechanics?

QIS Hardware and Apps

quantum operations

Qubit representation

Measurement

Overlap

Matrix multiplication

Qubit operations

Multiple Quantum Circuits

Entanglement

Deutsch Algorithm

Quantum Fourier transform and search algorithms

Hybrid quantumclassical algorithms

Quantum annealing, simulation and optimization of algorithms

Quantum machine learning algorithms

Exercise 33: Quantum operations multiexercises
Module 22: Introduction to quantum mechanics

Quantum mechanical theory

Wave function

Schrodinger's equation

Statistical interpretation

Probability

Standardization

Impulse

The uncertainty principle

Mathematical Tools of Quantum Mechanics

Hilbert space and wave functions

The linear vector space

Hilbert's space

Dimension and bases of a Vector Space

Integrable square functions: wave functions

Dirac notation

Operators

General definitions

Hermitian adjunct

Projection operators

Commutator algebra

Uncertainty relationship between two operators

Operator Functions

Inverse and Unitary Operators

Eigenvalues and Eigenvectors of an operator

Infinitesimal and finite unit transformations

Matrices and Wave Mechanics

Matrix mechanics

Wave Mechanics

Exercise 34: Quantum mechanics multiexercises
Module 23: Introduction to quantum error correction

Error correction

From reversible classical error correction to simple quantum error correction

The quantum error correction criterion

The distance of a quantum error correction code

Content of the quantum error correction criterion and the quantum Hamming bound criterion

Digitization of quantum noise

Classic linear codes

Calderbank, Shor and Steane codes

Stabilizer Quantum Error Correction Codes

Exercise 35: Noise Model, Repetition Code and quantum circuit
Module 24: Quantum Computing II

Quantum programming

Solution Providers

IBM Quantum Qiskit

Amazon Braket

PennyLane

Cirq

Quantum Development Kit (QDK)

Quantum clouds

Microsoft Quantum

Qiskit


Main Algorithms

Grover's algorithm

Deutsch–Jozsa algorithm

Fourier transform algorithm

Shor's algorithm


Quantum annealers

DWave implementation

Qiskit Implementation

Exercise 36: Quantum Circuits, Grover Algorithm Simulation, Fourier Transform and Shor
Module 25: Quantum Machine Learning

Quantum Machine Learning

Hybrid models

Quantum Principal Component Analysis

Q means vs. K means

Variational Quantum Classifiers

Variational quantum classifiers

Quantum Neural Network

Quantum Convolutional Neural Network

Quantum Long Short Memory LSTM


Quantum Support Vector Machine (QSVC)

Exercise 37: Quantum Support Vector Machine
Module 26: Tensor Networks for Machine Learning

What are tensor networks?

Quantum Entanglement

Tensor networks in machine learning

Tensor networks in unsupervised models

Tensor networks in SVM

Tensor networks in NN

NN tensioning

Application of tensor networks in credit scoring models

Exercise 38: Neural Network using tensor networks
Upstream Oil and Gas Models
Module 27: Geophysics and seismic data processing and interpretation using AI and Quantum AI
One of the most crucial tasks in seismic reflection imaging is to identify salt bodies with
with great precision. Traditionally, this is achieved by visually selecting the boundaries between salt and sediment, which requires a great deal of manual work and can introduce systematic bias. With the recent progress of deep learning algorithms and increasing computational power, great efforts have been made to replace human effort with machine power in the interpretation of salt bodies.
Convolutional neural networks (CNN) is revolutionizing the field of computer vision and CNNbased classification is demonstrated using a stateoftheart UNet network structure, along with the ResNet residual learning framework, to delineate the body of salt with high precision.

Seismic reflection experiment

Inverse problem: creating images and velocity models

Acoustic wave equation

The process of velocity estimation, and imaging of scatterers in the subsurface

Saline interpretation with machine learning

UNet for semantic segmentation

Loss function for UNet

Laminated KFold

Interpretation results

Classification of images in Real Estate


Problem Statement

Deep Learning Problem Formulation

Project and Data Source

Image Dataset

Evaluation Metric

Exploratory Data Analysis

Image Preprocessing

Model Training

Productionizing


Image Classification: Datadriven Approach

Convolutional Neural Networks

Data and the Loss preprocessing

Hyperparameters

Train/val/test splits

L1/L2 distances

Hyperparameter search

Optimization: Stochastic Gradient Descent

Optimization landscapes

Black Box Landscapes

Local search

Learning rate


Weight initialization

Batch normalization

Regularization (L2/dropout)

Loss functions

Gradient checks

Sanity checks

Hyperparameter optimization

Architectures

Convolution / Pooling Layers

Spatial arrangement

Layer patterns

Layer sizing patterns

AlexNet/ZFNet/Densenet/VGGNet/UNet/Resnet


Convolutional Neural Networks tSNE embeddings

Deconvnets

Data gradients

Fooling ConvNets

Human comparisons Transfer Learning and

Finetuning Convolutional Neural Networks

Performance Metrics

Accuracy

F1Score

AUCROC

Cohen Kappa Coefficient


Exercise 29: Deep CNNBased Architecture for Automatic Interpretation of Salt Bodies from Seismic Data
Module 28: Production engineering using machine learning
Production engineering encompasses various activities, such as measuring, analyzing, modeling, prioritizing and executing actions to improve productivity and profitability. Production engineers generally consider multiple options to improve the conditions of the production system from a technical, economic and environmental point of view. With the introduction of ML in the oil and gas industry, numerous applications have been published to assist in the research of flow state identification, performance measurement, and identification of production capacity problems and limitations.

Production engineering

Potential Well Rate Prediction

Increased well knowledge with virtual sensors

Virtual rate measurement

Predicting well rates from indirect measurements

Predicting well rates for gaslift wells

Well failure prediction

Data Feature Engineering

Root Cause Analysis (RCA)

Actions to improve performance

Predicting poor well performance

Water cut management based on speed

Critical Oil Rate Prediction

Optimal well spacing in naturally fractured reservoirs NFRs

Reservoir threshold radius prediction

Exercise 40: Critical Oil Rate Model using linear regression, random forest regression and quantum neural networks

Exercise 41: Classical neural network model and quantum neural networks of Rate and Well Pressure

Exercise 42: Well Failure Clustering Using KMeans

Exercise 43: Well Rate Model of the random forest and Gradient Boosting Machines
AI and Quantum Computing for Reservoir Engineering
Module 29: Machine Learning in Reservoir Engineering
Reservoir engineering is an interdisciplinary field that integrates mechanics, geology, physics, mathematics and computer science as research tools to economically recover hydrocarbon resources in underground formations.
With the improvement of computing architecture, ML can also efficiently extract nonderived information from the explosion of realworld data collected in oil industry applications. As a powerful datadriven technique, ML is being widely accepted to help and improve our understanding of drilling, production and reservoir areas. In particular, ML has been widely used in reservoir engineering and has achieved excellent results, such as prediction of permeability, porosity and tortuosity, prediction of shale gas production, reservoir characterization, reconstruction 3D digital core analysis, well test interpretation, rapid shale gas production optimization, well log processing and historical matching.

XGBoost for cumulative production forecasting

Random Forest for the identification of water invasion patterns; property estimation

SVM for surface oil rate prediction; permeability estimation

Gaussian Regression Process for early oil and gas production

ARIMA for production dynamics

LSTM for production dynamics; identification of reservoir models

CNN for well test interpretation and production dynamics

Exercise 44: Forecasting Future Production Rates of an Oil Well Using Decline Curve Analysis

Exercise 45: Forecasting future production rates of an oil well using ARIMA

Exercise 46: Production dynamics of an oil well using LSTM

Exercise 47: Production dynamics of an oil well using Bayesian LSTM
Module 30: Advanced Oil Production Forecasting Models with Multivariate Data
Accurate forecasting of well production is crucial for petroleum engineers to identify rapid declines in oil production and select appropriate recovery mechanisms. It also helps estimate the ultimate recovery and useful life of the well, leading to sustainable development and balancing economic inputs and outputs. However, predicting oil production accurately can be challenging due to the complexity of subsurface conditions. Heterogeneous formation properties, multiphase fluid flow, uncertain subsurface conditions, and frequent in situ operations can significantly impact production, making forecasting models more complex. Therefore, advanced forecasting models are presented.

The oil production forecasting model

Multivariate data analysis

raw data import

oil volume forecast

evaluating the predictive performance of the model

comparison with reference models.

Advanced Models

TCNTemporal Convolutional Networks

LSTMLong Short Term Memory

RNNRecurrent Neural Network

GRUGated recurring units

QLSTMQuantum Long Short Memory

Advanced model with genetic algorithms

Exercise 48: LSTMLong Short Term Memory for oil production forecasting

Exercise 49: GRUGated recurring oil production forecasting units

Exercise 50: Quantum Long Short Term Memory for oil production forecasting

Exercise 51: TCNTemporal Convolutional Networks for oil production forecasting

Exercise 52: Genetic algorithm model for oil production forecasting
Probabilistic Machine Learning for Oil and Gas
Module 31: Geomodeling used Gaussian process regression
Geomodeling is an important step in the exploration and production planning process. Geomodeling is carried out after understanding the structural framework of the reservoir based on the information extracted during the seismic interpretation. Machine learning can help us estimate petrophysical property values on a threedimensional mesh. Variograms are the basis of almost all 3D spatial modeling techniques of industrial relevance.

Variogram

Data Description

Basic Well Log Data Analysis

Gaussian process regression

Formulation

Radial Basis Function (RBF) Kernel

Exercise 53: Geomodeling using Gaussian process regression
GENERATIVE AI IN OIL & GAS UPSTREAM
Module 41: Generative AI

Introducing generative AI

What is Generative AI?

Generative AI Models

Generative Pre trained Transformer (GPT)

Llama 2

PaLM2

DALLE


Text generation, Image generation, Music generation, Video generation

Generating text

Generating Code

Ability to solve logic problems

Generating Music

Enterprise Use Cases for Generative AI

Overview of Large Language Models (LLMs)

Transformer Architecture

Types of LLMs

OpenSource vs. Commercial LLMs

Key Concepts of LLMs


Prompts

Tokens

Embeddings

Model configuration

Prompt Engineering

Model adaptation

Emergent Behavior

Specifying multiple Dataframes to ChatGPT

Debugging ChatGPT’s code
Human errors 
Exercise 54: Embeddings for words, sentences, question answers

Exercise 55: Embedding Visualization

Exercise 56: PCA (Principal Component Analysis)

Exercise 57: Embeddings on Large Dataset

Exercise 58: Prompt engineering

Exercise 59: Advanced Prompting Techniques

Exercise 60: Large Language Models (LLMs)

Exercise 61: Retrieval Augmented Generation

Exercise 62: Traditional KMeans to LLM powered KMeans

Exercise 63: generative AI in Reservoir Engineering

Exercise 64: Transformers Oil Production Forecasting with generative AI