AM-SciML vs degrees in CS/Data Science/Computational Math

AM-SciML focuses on Machine Learning models and scientific computing methods, targeting applications in Science, Medicine, and Engineering.

The first main difference that sets AM-SciML apart from a CS degree is the focus of AM-SciML on Science and Engineering applications, with exposure to topics from the physical sciences, the life sciences, differential equation models, and control engineering.

The second main difference with a CS degree is the emphasis in AM-SciML on the mathematical foundations that underly Scientific Machine Learning models. AM-SciML builds on foundational courses in computational mathematics, calculus, linear algebra, scientific computing, differential equations, and optimization, and has less emphasis on coding and CS areas such as operating systems and databases. CS is a great degree for students interested in the broader field of Artificial Intelligence, beyond Scientific Machine Learning.

Computational models studied in AM-SciML often combine first-principle physical models (e.g., differential equation models) with Machine Learning approaches. As such, Scientific Machine Learning models learn from data while incorporating the rules of physics, biology, and engineering. This makes them more accurate, interpretable, and efficient for solving real-world problems in those areas.

AM-SciML provides a broad set of specialized courses in Scientific Machine Learning, neural networks, mathematics of deep learning, statistical learning, and data-driven mathematical modeling, facilitating the application of Machine Learning techniques to broad areas of Science, Medicine, and Technology.

In terms of methodology, AM-SciML focuses on Machine Learning models and scientific computing methods, targeting applications in Science, Medicine, and Engineering. Whereas in AM-SciML you will learn about mathematical and computational models driven by data, Data Science degrees often focus more on statistical models and CS aspects of Data Science such as databases and data engineering, including the use of these techniques for Business applications.  

Computational models studied in AM-SciML often combine first-principle physical models (e.g., differential equation models) with Machine Learning approaches. As such, Scientific Machine Learning models learn from data while incorporating the rules of physics, biology, and engineering. This makes them more accurate, interpretable, and efficient for solving real-world problems in those areas.

The AMATH courses in AM-SciML provide important mathematical background for understanding Scientific Machine Learning models and offer students a choice of courses that provide a mathematical and computational introduction to topics from the physical sciences, the life sciences, differential equation models, and control engineering, thus facilitating the application of Machine Learning techniques to broad areas of Science, Medicine, and Technology.

Compared to Waterloo’s Computational Mathematics (CM) program, AM-SciML has a clearer focus on Machine Learning, combined with first-principle physical models from Applied Mathematics, and has a clearer focus on applications in Science and Engineering. The AMATH courses in AM-SciML provide important mathematical background for understanding Scientific Machine Learning models and offer students a choice of courses that provide a mathematical and computational introduction to topics from the physical sciences, the life sciences, differential equation models, and control engineering, thus facilitating the application of Machine Learning techniques to broad areas of Science, Medicine, and Technology. (Note that the CM program only requires one AMATH course.)