SYS 5581

I took this course on a whim last fall and it was one of the best decisions I've made. I'm a Systems' student and that was at the beginning of my fourth year; I was hoping to get a bit more serious in preparation for graduate studies and this course could not have been more perfect for that. I gained a much deeper understanding of probability theory while also getting a practical understanding of how to use it to model forecast-decision systems.

Professor Krzysztofowicz is extremely knowledgable, this is the only course he teaches in the fall, and it covers theories that he's worked on over the past 30+ years. As such, he has a lot of passion for the content and is very capable of offering one-on-one guidance. However, he has a very traditional teaching style: lectures were not recorded and were handwritten on the chalk board, assignments were to be submitted by hand rather than online, there was an expectation that notes would be taken by hand, etc. For these reasons, you should stay away if you're just looking for an easy course at the end of your undergraduate degree. However, if you're looking for a transition into higher-level material with a professor who can offer a lot of help if you begin to struggle, then I couldn't recommend this course more.

Some of the reviews on here paint a very different picture of Professor Krzysyztofowicz than what I've experienced, so I wanted to address a couple of the things people have mentioned. All of his reviews are for APMA 3100 and SYS 3060, both of which are required courses, which limits the amount of control the professor has over the content and assignments. Since this course is an elective, he has full control over these things, and the result is that the assignments followed directly from the content covered, they were graded very quickly, and they could be corrected for an improved grade if you didn't like your original score. They were fairly difficult, but not unreasonably so considering you are generally given 2+ weeks to complete them. There were also no exams, just the homeworks and a 30-minute mini-lecture at the end of the course.

As a side note, Professor Krzystofowicz teaches another course in the spring that I have also taken. It details some very flexible (non-linear, non-Gaussian, heteroskedastic) forecasting models that can be used within the broader forecast-decision models presented in this course. I just think it's useful to know, as they make for a very nice 2-course sequence.