ECS 189G, Winter 2020

Norm Matloff, matloff@cs.ucdavis.edu

[Informative, intended to help you do well in the course. NOT intended to intimidate. :-) ]

Course Prerequites and Content

Please carefully read the course flyer, especially regarding prerequisites -- they matter a lot!

We will use the R programming language. It is NOT assumed that you have prior background in R. However, you are expected to pick up the basic material on your own. It is required that you read my quick online tutorial. It's designed for nonprogrammers, so you'll find it quite easy to follow.

I Teach Differently from Others

• Quizzes EVERY week. Open book.
• End-of-quarter term project, with WRITTEN report.
• Group work plays a central role in homework, term project. Constant coordination with teammates is key.
• Quizzes are taken on your laptop computer, in class, using the OMSI system.
• Quizzes emphasize insight, not rote memorization of formulas, patterns or programming language syntax.

Good Fit for You?

This is a course on recommender systems, a branch of machine learning. Successful application of ML methods requires keen insight and intuition; it is not cookbook "Step A, Step B,..." methodology.

I believe you will enjoy and value this course if:

• You feel that exploring the How? and the Why? of a course is an integral part of a university education. You thus appreciate digressions in lecture, provided they are related to the class.
• You are comfortable with being required to do detailed reading and being tested on it.
• You believe that a student's hard work should be rewarded with a generous course grade. (See below.)

• You plan to take it for a P/NP grade.
• You feel that all possible aspects of a course should be covered in lecture, rather than some being relegated to the assigned reading.
• You prefer a highly structured class that sticks to fundamentals, has practice exams that are similar to the real exams, provides detailed outlines to programming assignments, and so on.

Exams

• No midterms or final.
• Group term project in lieu of final, due on the day of the scheduled final. Make sure this fits into your anticipated final exam schedule in terms of workload.
• Quizzes are given in discussion section EVERY WEEK. Quiz problems range from straightforward to very challenging.
• There is a special Group Quiz on the last day of lecture.
• Quiz problems will often refer to specific pages in the book. (Open book Quizzes, no memorization needed.) It is REQUIRED that you read the textbook, in complete detail, and give it careful thought.
• Quiz scores tend to be low, but letter grades are generous.
• All old Quizzes may be viewed here.
• Quizzes form 60% of your course grade.

Homework

• About 3-4 assignments during the quarter.
• Homework is done in GROUPS.
• Grading is done interactively, by Group, but INDIVIDUALLY by Group member.
• TA will ask each member of the Group questions, both about the assignment and about the course material in general. Each Group member receives a separate grade; one member may get an A+ while another gets a D.

Textbooks:

Our main textbook is here. But PLEASE NOTE: The book is short a few chapters. It was developed when I taught the course in Fall 2018, when the quarter was substantially shortened due to regional fires. I will be adding to the book as the quarter progresses.

As noted above, you are required to know/learn R, to the degree presented in the above-mentioned tutorial and the appendix in our main textbook. You will be using R heavily in the quizzes. Ours is not a programming course, and the level of programming used usually will be simple, but it IS used a lot.

You are REQUIRED to have hard copies of these documents -- actual paper, not electronic (best to go to a copying store). Please print them BEFORE classes start. Quizzes are open-book/open-notes.

Grading

• Quizzes 60%, homework/term project 40%.
• Lowest 2 quizzes dropped.
• Huge bonus for a good term project.
• Through various "special offers," many students find that their course grade ends up much higher than their average grade in the quizzes.

Example: In ECS 132 last quarter, there was a student with a C average on the quizzes, and a B average on the homework. His team had an A- on the term project. This student ended up with an A course grade! Again, this was due to the "special offers." This was not an isolated incident at all; there were many students whose course grade was much higher than their quiz grades.

Lecture/Learning Format

• It is expected that the student be NON-PASSIVE in learning. Most of the student learning comes from reading the textbook in very careful, thoughtful detail.
• I used to lecture in the traditional way -- I spent time writing on the board, the students spent time copying it to paper.
• Then I realized, "What a colossal waste!" So, now I give you the printed notes, so you and I can spend the time discussing the material.
• So, in lecture, we DISCUSS the book. I almost never write on the board, nor do I have Powerpoint slides.
• It is NOT recommended that students read the book ahead of the lecture. In the lecture, I prepare you to do the reading.
• In some cases, I will not lecture on a certain section of the book, and simply have you read it on your. I choose for such sections ones that I believe you'll have no trouble with. But if you do have questions, please make sure you ask me or the TA.
• Goals of the lecture:
  • Clarify issues brought up by students concerning reading related to the last lecture.
  • Prepare students for the next reading material, by giving an overview of what the material will do, and going into the more difficult examples in detail .

• You could decide to skip lecture and just read the book, but it IS really helpful to go to lecture. There will be spur-of-the-moment examples and insights, answers to students' questions. etc. And you are responsible, in Quizzes and interactive Homework grading, for everything that comes up in lecture.