About Prob 140

A. Adhikari

What is Prob 140?

Prob 140 (formally Statistics 140 or STAT 140) is a probability course for undergraduates who have taken Data 8, have a math background, and wish to go deeper into the theory of data science.

The emphasis on simulation and the bootstrap in Data 8 gives students a concrete sense of randomness and sampling variability. Prob 140 will capitalize on this. Because of the students’ backgrounds, Prob 140 will move swiftly over basics, avoid approximations that are unnecessary when SciPy is at hand, and replace some of the routine calculus by symbolic math done in SymPy. This will create time to focus on the more demanding concepts that are part of the theoretical foundations of data science.

Prob 140 contents have been selected based on consultation with faculty who regularly teach Stat and CS courses in advanced statistical topics including machine learning. Specifically, Prob 140 is intended to provide good preparation for CS 189 and Stat 154, with the goal that students should be able to move from Prob 140 to a machine learning class without necessarily having to take a semester of theoretical statistics in between.

The class requires more mathematics than Stat 134, the Statistics department’s standard course in classical probability theory; in addition to calculus, Prob 140 requires linear algebra. Computational power in Prob 140 allows students to solve problems that are intractable by other methods. Students also explore the standard mathematical theory graphically and by simulation, and thus develop a more firm grasp of the concepts than they might by using math alone.

What requirements does Prob 140 satisfy?

Prob 140 satisfies the probability requirement for the proposed new Data Science major. The FAQ page of the Data Science Education Program website has this to say:

• Should I be interested in STAT 140?

  • STAT 140 covers Probability for Data Science. It uses computation to develop the theoretical foundations of data science. It satisfies the STAT 134 prerequisite for advanced statistics courses as well as for the Statistics major. It is also a good choice for students who may be thinking of majoring in Data Science. STAT 140 has Data 8 and 1 year of calculus as prerequisites; starting in Spring 2018, Math 54 or EE16A will be a co-requisite.

For the Statistics major and minor, and for Statistics courses numbered 135 and above, Prob 140 satisfies the same requirements as Stat 134 does. If a Statistics course currently requires Stat 134, then Prob 140 will fulfill that requirement too. A letter grade of B- or better in Prob 140 will satisfy the corresponding Stat 134 grade requirement for entry into the Statistics major.

Prob 140 satisfies Technical Elective requirements for some non-Stat majors, and students can petition others. Please direct your inquiries to the other major and include the link to the Prob 140 website if needed.

Who can take Prob 140?

Prob 140 is restricted to undergraduates who:

• Have not taken Stat 134; only one of Stat 134 and Prob 140 can be taken for credit
• Have had a year of calculus at the level of Math 1A-1B and preferably higher; Prob 140 involves some double integration and partial derivatives
• Have taken or are concurrently taking linear algebra in Math 54 or EE 16A or Math 110, or have taken an equivalent linear algebra course at another college
• Have a passing grade in Data 8 (P is fine if the course was taken P/NP); the EdX MOOC Data 8X does not satisfy this requirement

Fall 2018 Enrollment Note: Starting in April 2018, Prob 140 will be part of a pilot program in which CalCentral will check prerequisites automatically. As I understand it, students who satisfy the prerequisites above will be able to enroll in the usual way if there is space. Others will be placed on the waiting list.

If there is space in the class after the Spring enrollment period, we will enroll waitlisted undergraduates who have not taken Data 8 but have passing grades in

• DS 100, or
• Both Stat 20 (or Stat 21 or Stat 131A) and CS 61A

in order of waiting list position until the class is full.

These alternatives to Data 8’s inference and programming curriculum are due to a Data 8 “grandfathering” period during the upcoming transition as the Data Science major becomes available. Students who satisfy the inference requirement in these ways will be responsible for studying the necessary Data 8 material on their own; it will not be retaught in Prob 140.

All of these requirements will be enforced as they have been in past semesters. I realize that many students have taken more advanced computing or prob/stat courses than required, but not the ones required. However, they are not the intended audience of Prob 140.

What is the best way to succeed in Prob 140?

Prob 140 is an upper division course that is primarily mathematical. Students are expected to have a corresponding degree of independence, discipline, and mathematical maturity.

In every class I teach, I aim to make learning as efficient and enjoyable as possible. In particular, I try to make the lectures a great way to learn the material, so show up (there will be no videos) and ask lots of questions. A preparation guide is provided each week; it lists the reading and practice that students should do before attempting the week’s assignments. These lists aren’t just decoration. They are designed to be helpful. Homework will go faster if you’ve done the reading and some practice first.

But rather than reading more advice from me, listen to your fellow students instead. Here are quotes from student evaluations of the pilot course. They are from answers to the question, “What advice would you give to another student who is considering taking this course?”

• [E]ach problem on the homework and lab is meant to teach you a new use of the techniques developed in class.
• Review the prereq worksheet before the beginning of the semester … Also, refrain from starting the online homework right away. Read the corresponding sections and do the practice problems first. Overall, this class requires a lot of time and practice, but honest effort will be rewarded.
  • be sure to invest enough time, especially homework, they’re very good guidelines.
  • go to office hours, talk to professor Adhikari!
  • review the material right after class when you still remember. it helps tremendously with hw. Just read the notes.
  • do more questions! it’s still very much like 134 where you should basically go through a lot of questions.
• Amazing class! Definitely take over 134 because it’s a modern way to learn probability.
• Be prepared to work your ass off, but this class is worth it!

Will Prob 140 be offered in both Fall 2018 and Spring 2019?

Yes.