Machine learning is a convergence of linear algebra, statistics, optimization, and computational methods to allow computers to make decisions and take action from data. Examples of machine learning are now pervasive and are expected to further influence transportation, entertainment, retail, and energy industries. This engineering course reviews theory and applications of machine learning to engineering applications with a survey of unsupervised and supervised learning methods.
The course combines mathematical details with several case studies that provide an intuition for machine learning with methods for classification, regression, and dimensionality reduction. A second phase of the course is a hands-on group project. The engineering problems and theory will guide the student towards a working fluency in state-of-the-art methods implemented in Python.
- Visualize data to understand relationships and assess data quality
- Apply linear algebra, statistics, and optimization techniques to create machine learning algorithms
- Understand engineering and business objectives to plan applications
- Assess data information content and predictive capability
- Detect overfitting and implement strategies to improve prediction
- Master the use of machine learning packages with understanding of how hyperparameters can be adjusted to improve performance
- Understand the differences between classification, regression, and clustering and when each can be applied
- Communicate machine learning decisions with uncertainty quantification
- Implement machine learning techniques successfully to complete a group project