Prof. Norman Matloff, UC Davis
I am deeply committed to teaching, and have even been fortunate to win a couple of teaching awards, notably the campus-wide Distinguished Teaching Award (scroll down at that URL).
However, many of my methods are quite different from those of many of your instructors. The purpose of this document is to give you an accurate idea of what my ECS 132 class will involve. Which of the rumors are false, and which actually contain a grain of truth? :-)
Is ECS 132 a math class or a programming course?
It's a math course, similar in content to STA 131A and MAT 135A. But unlike those courses, ECS 132 uses programming to supplement the math.
What programming language is used?
We use R. Python is popular for machine learning, but for general data science R is far more widely used. As I like to say, "R is written BY data scientists, FOR data scientists." You may find my essay, R vs. Python for Data Science, of interest.
I have a quick tutorial that would be good for you to go through before the quarter starts. It's been used successfully by hundreds of people all over the world (700+ GitHub stars), most of them non-CS, so you as a CS major should be able to breeze through it.
BTW, from a pure CS point of view, R has a lot of interesting features, such as for metaprogramming ("program code writing program code").
How important are the prerequisite courses?
VERY important.
-
Calculus: We use derivatives, integrals and infinite series. Specific methods, say, integration by partial fractions, are less important than an intuitive understanding, e.g. that a derivative is the slope of a curve, the second derivative shows the concavity (direction and degree) of a curve, and an integral is the area under a curve.
-
Linear algebra: Matrix multiplication, matrix inverse and transpose operations, the notion of a linear combination of vectors etc. will all be used a lot. We will use eigenvalues somewhat. We won't do much with vector spaces, bases, etc., though BTW they are crucial to more advanced data science.
See further details regarding our usage of math here.
Is it true that NM lectures right out of the textbook?
Yes. The book evolved originally from my course notes, so this is what I discuss in class. So, you can focus on what I'm saying about the concepts, rather being distracted by taking notes.
Does that mean we need not go to class?
No, class attendance is vital to doing well in the course. During lecture I elaborate on what is in the book. Note that many of these elaborations will show up later on the quizzes.
If the book = lecture notes, how should we approach NM's lectures?
Since "the notes have already been taken for you" (in the form of the book), you will take rather few notes. Your time in lecture should be spent listening carefully to the lecture--and even more carefully to answers thtat I give to student questions during lecture--and only take notes on points that arise that are not in the book.
Is it true that NM does not go over examples of the concepts in class?
Absolutely false. There are examples on almost every page in the book, and I cover them. Often I will also discuss how things might change if conditions of an example were changed. Again, often this material on "what if" will show up later on the quizzes.
Does NM sometimes skip a section in the book during lecture, asking the students to read that section on their own?
Yes. These are sections that you should be able to delve into on your own, if you've been keeping up with the material. Of course, the TAs and I are always happy to answer any questions you may have.
What is NM's grading like in this course?
As you know, in the last few years, a few instructors have been giving extremely liberal grades. I am not one of them, but I think my grading is reasonable. Here are my averages since Fall 2019:
mean course grade given: 3.25 (B+)
proportion of A+ grades: 17%
That mean-grade figure places me about in the middle of all instructors of the course.
The A+ figure is somewhat higher than average. It stems largely from extra incentives I give at the end of the course for high-quality term projects.
What are the course grades based on?
Weekly quizzes; group homework with individual interactive grading; term project. Your two lowest (letter) quiz grades are thrown out.
Is it true that NM's practice tests are nothing like the real quizzes?
Frankly, I don't believe in practice tests. A University of California course should not center on rote memorization of patterns and formulas. I hope you agree that getting a good score on an exam that closely follows a practice test is not something you should be proud of.
I aim for you to be able to USE the material, not just get a grade on your transcript.
The quizzes test insight, but the cutoffs for letter grades are set very liberally; a typical quiz will have a cutoff of 70 for an A, 50 for a B and so on. Again, if you have been keeping up with the material, you will do fine.
I do publicly make available all the past quizzes I've ever given, with solutions. They are not practice tests, but they certainly will give you a good idea as to the general nature of what I ask.
Note that all quizzes are open-book, open notes. You take the quizzes on your laptops, in class, using my OMSI system.
How does the interactive homework grading work?
You work on the assignments in groups of 3 or 4 students, and submit as a group. For grading, you make an appointment as a group, but the TA will ask each of you questions individually, such as "What is going on in this step of your solution?" and "What if the prompt for the problem had been different, such-and-such? How would your solution change?" Your grade on the assignment is a combination of the quality of the group written submission, and your answers to the individual questions.
Does this require a lot of reading?
It's required that you read the textbook in detail, and quiz questions will often refer to a specific page in the book. The book is about 200 pages long, so you will be reading, say, 4 or 5 pages per day.