Probably the most important class that you’ll take as anyone working with data.

What is ECE 435?

The course name Machine Learning and Pattern Recognition is a very vague notion. In concrete terms, this course is a graduate-level (no really, it’s cross listed as ECE 535 and majority graduate students) proof-based data-driven machine learning course.

When Should You Take It?

As soon as possible, as long as you have a solid grasp of linear algebra, multivariable calculus, and probability (whether or not you have actually taken those courses). If you want to take courses, I would recommend MAT 217, MAT 201, ORF 309 for requisite knowledge, but that might be overkill. I say take this course as soon as possible because a solid mathematical foundation of data-driven machine learning algorithms will make all the concepts much, much more clear and allow you to apply these skills anywhere.

You may think that you understand ML from a COS 324-style hand-wavy machine learning course, but I didn’t, and you don’t.

Additional Resources

• For linear algebra and visualization: 3Blue1Brown’s Linear Algebra series
• For Singular Value Decomposition, Principal Component Analysis, and Linear Regression: Eigensteve’s SVD Playlist
• For Gaussian Mixture Models and Expectation Maximization, Ritvikmath’s videos
• For Maximum Likelihood Expectation: see StatQuest

My Advice and Notes

To Be Added post-completion