Statistics 154
Modern Statistical Prediction
and Machine Learning
Spring 2022
Instructor: Nusrat Rabbee,
rabbee@berkeley.edu.
Office:
305 Evans.
Office
hours: F 2:00-4:00pm (zoom), or after class or by appointment
Graduate
Student Instructor: Austin Zane, austin.zane@berkeley.edu
Office hours: W
2-4p (428 Evans/zoom) ; Th 3-5p (428 Evans)
Website: http://www.stat.berkeley.edu/users/rabbee/s154.
We will post announcements, assignments, lecture notes etc. on
bcourses.berkeley.edu. Check regularly for updates.
Schedule: There will be lectures two days a week, TTh 5:00-6:30, in Etcheverry
3108. There will also be weekly sections, scheduled M 9-11p or 3-5p, starting
1/24. Attendance to both lectures and sections is highly encouraged.
Textbooks:
1. Required: James,
Witten, Hastie, Tibshirani. An Introduction to Statistical Learning. Hardcopy.
Online
version. (Courtesy of
the authors)
2. Optional: Hastie, Tibshirani and Friedman. The Elements of Statistical Learning. Second
Edition. This book is more mathematically advanced than the one above. Hardcopy.
Online
version. (Courtesy of the authors). This text will not be used directly for
this course and is a reference for more theoretical details.
3. Optional: Rabbee, N. Biomarker analysis in
Clinical Trials Using R. This book is mathematically suited for advanced
studies than the first one above. It may
be used as a reference for graphical models for visualization methods. Copies
of relevant chapters will be provided (courtesy of author).
Exams
and grading: There will
be two written exams (Th during class or take home) and a final project
(due M 05/09 @noon). There will be three to four quizzes during section. There
will be no make-up quiz, written exam or final project due date
adjustments; do not take the class if you are not available at these dates and
times. Your grade will be 30% best three quizzes, 30% written exams, 40% final
project.
Assignments: There will be six to seven assignments. They are
announced in bCourses on Fridays. The assignments are
not to be handed in. You should do the assignments in teams or by yourself. The
quizzes will be similar to the assignment sets.
Academic
Integrity: You may collaborate with your assigned
team for the final project and you will share the same score. No collaboration
is allowed in the quizzes or exams. Penalties for cheating will be severe. Here
are more details.
Special instructions for final project teams will be announced later.
Communicating: Questions about lectures should be directed primarily
to me after lecture or in office hours, about section and assignments primarily
to the GSI. Emails are generally discouraged. Write to me only if you have any
pressing administrative issues. Emails should be brief, marked stat 154 in
the subject and crisp for a good chance at being answered. Regardless, you are
encouraged to come to any of our office hours or stay after class: talking
is usually more effective than sending email. Feedback is always welcome.
Support: If you experience stress or challenges,
depression or anxiety, you are strongly encouraged to seek help. Please ask department
staff, faculty or a trusted family member - for support sooner rather than
later in the semester. We are here to help you get connected to the support you
need.
Prerequisites: Mathematics 53 and 54 or
equivalents; 110 is highly recommended. Statistics 135 or equivalent.
Statistics 133 preferred. Stat 151A is recommended. Scripting language and R
experience required. Mathematics 55 or equivalent exposure to counting
arguments is recommended but not required.