Use Cases of Machine Learning for Business, Engineering, and Science

Case 1: Feature Engineering in Li-Ion Battery Life Prediction

Introduction to Feature Engineering
Predicting Lithium Ion Battery Lifetime
Identifying Features
Three Types of Regularization
Applying Feature Engineering
Why Feature Engineering and Elastic Net Matters
Graded assignment

Case 2: Machine Learning for Computational Imaging

Case 3: Seismic Deepfakes: Neural Nets to Generate Missing Data

Case 4: Prediction of Oil and Gas Production

Case 5: Machine Learning in Geometric Representations

Case 6: Quantifying Risk in Complex Systems Using Machine Learning

Case 7: Machine Learning for Accelerating Computational Materials Discovery

Case 8: Practical Machine Learning in Composite Design

Case 9: Machine Learning for Data Assimilation and Inverse Problems

Inverse Problems from Energy to the Geosciences
Uncertainty Quantification for Inverse Problems: The Bayesian Approach
Computational Challenges of Bayesian Inversion
Video from Finnos Oy
Dimensionality Reduction
Surrogate Modeling
Data Assimilation: From Static to Sequential Data
Nonlinear Ensemble Methods for Data Assimilation
Applications of Machine Learning in the Aerospace Industry
Predictive Health Monitoring of Aerospace System
Graded Assignment
Publications

Conclusion

Course Content

Case 1: Feature Engineering in Li-Ion Battery Life Prediction

Introduction to Feature Engineering

Predicting Lithium Ion Battery Lifetime

Identifying Features

Three Types of Regularization

Applying Feature Engineering

Why Feature Engineering and Elastic Net Matters

Graded assignment

Case 2: Machine Learning for Computational Imaging

Machine Learning for Computational Imaging

Inverse Problems

Phase Retrieval

Phase Extracting Neural Network

Perceptual Loss

Learning to Synthesize

Tomography

Computational Imaging Conclusion

Graded Assignment

Case 3: Seismic Deepfakes: Neural Nets to Generate Missing Data

Introduction to Seismic Waves

Wave Equations

Inversion

From High to Low Frequencies

Where Neural Networks Finally Come In

Training and Testing

Inversion Revisited

Publications

Graded Assignment

Case 4: Prediction of Oil and Gas Production

Minimizing Decision-Making Risk

Oil and Gas Leases: Predicting Future Production Rates

Linear Regression and Predicting with Data

Predicting Future Production Rates: Results

Graded Assignment

Case 5: Machine Learning in Geometric Representations

Introduction to Machine Learning from Geometric Representations

Two Modalities of 3D Geometric Data

Deep Learning from Point Clouds

Applications of Point Cloud Learning

Deep Learning from Vector Data

Frontiers in 3D Learning

Graded Assignment

Case 6: Quantifying Risk in Complex Systems Using Machine Learning

Quantifying Risk of Extreme Events

Probabilistic Description of Extreme Events

Challenges Related to Extreme Events

A Better Approach to Calculating Probability

Active Learning and Optimal Experimental Design

Using a New Output-Weighted Criterion

Q Criterion Results

Graded Assignment

Case 7: Machine Learning for Accelerating Computational Materials Discovery

Introduction to Inorganic Chemistry

The Machine Learning Tradeoff

Representations for Transition Metals

Training Set Performance and Details

Feature Selection

Uncertainty Quantification

Accelerating Discovery

Graded Assignment

Case 8: Practical Machine Learning in Composite Design

What Is Materials Science and Engineering?

Introducing Machine Learning to Materials Design

Using Machine Learning for Image Classification

Complementing Conventional Methods with Machine Learning

Predicting Fracture Propagation

Hype vs. Reality: Challenges for the Future

Graded Assignment

Case 9: Machine Learning for Data Assimilation and Inverse Problems

Inverse Problems from Energy to the Geosciences

Uncertainty Quantification for Inverse Problems: The Bayesian Approach

Computational Challenges of Bayesian Inversion

Video from Finnos Oy

Dimensionality Reduction

Surrogate Modeling

Data Assimilation: From Static to Sequential Data

Nonlinear Ensemble Methods for Data Assimilation

Applications of Machine Learning in the Aerospace Industry

Predictive Health Monitoring of Aerospace System