Cogs 118D: Mathematical Statistics for Behavioral Data Analysis (Winter 2017)

Lectures: Tue/Thr 12:30-1:50, Mandeville B104
Sections: Wed 4:00-4:50, WLH 2115

Course website: http://thiscourse.com/ucsd/cogs118d/wi17/

Recent Announcements (more)

Displayed course content is preliminary; details may change prior to the start of the winter quarter.
(Thu Oct 20, 4:39 p.m.)

Instructors

Angela Yu (Professor)

/ phone: 858-822-3317

cogsci.ucsd.edu/~ajyu

Office Hours: M 4-5, Th 2-3 in SSRB 246

Darlene Guo (TA)

Office Hours: M 10-11, Tu 10-11 in SSRB 245

Course Description

This course will cover both traditional and contemporary statistical methods and apply them to real-life situations. Some distinctive features of the course:
- both frequentist and Bayesian methods (only course at UCSD to cover Bayesian stats)
- application oriented and one-quarter only (in contrast to Math 181A-181B-185)
- hypothesis-driven rather than exploratory (in contrast to Math 189)
- greater mathematical sophistication than introductory classes (Cogs 14, Psych 60, Math 11, Soci 60, Poli 30)

In contrast to the official pre-reqs (which will be changed for next year), students will be admitted to the class if they have taken linear algebra (Math 20F or 18), probability (Math 180A), and a programming class (Cogs 108 or 109, CSE 8B or 11, or can demonstrate programming competence by other means). All HW assignments will be in Matlab, there will be a Matlab tutorial in the first week for those who need an intro/refresher.

We will be using Piazza for course-related Q&A (counts toward your participation grade), announcements, and other interactions. Please only use email for urgent, personal issues that need immediate attention. Piazza sign-up: piazza.com/ucsd/winter2017/cogs118d . If you have any problems or feedback for the developers, email team@piazza.com.

Required textbook: "Probability and Statistics" by Degroot & Schervish (2012), 4th Ed.

The course covers Ch. 7-11 of the textbook. Students should have mastery of the probability concepts in Ch. 1-6 (covered in Math 180A) prior to the start of the course!

Readings

Additional readings, as needed, will be posted on the course calendar below.

Course Details and Policies

Grading

  • 20% participation
  • 14% demos
  • 25% HW assignments
  • 15% midterm exam
  • 25% final project
  • 1 % self-test (HW 1)
Extra credit: SONA participation

UCSD Policy on Integrity of Scholarship

Course Schedule

Week Date Topic Readings (by this date) Assignments due Notes
1 Tue Jan 10 Introduction
Review of probability
1.1-1.6, 1.10, 2.1-2.3, 3.1-3.8, 4.1-4.3, 4.5-4.7, 5.1, 5.6. 5.8, 6.1-6.3


Thu Jan 12 Review of probability (cont')
Review of Matlab (in section)
1.1-1.6, 1.10, 2.1-2.3, 3.1-3.8, 4.1-4.3, 4.5-4.7, 5.1, 5.6. 5.8, 6.1-6.3


2 Tue Jan 17 Conditional distribution, covariance, correlation
3.6, 4.6


Thu Jan 19 Estimation: Bayesian
7.1 - 7.3
HW 1

3 Tue Jan 24 Estimation: Frequentist

7.4-7.5



Thu Jan 26 Sampling distribution of statistics
7.7, 8.1, 8.2 HW 2

4 Tue Jan 31 Sampling distribution of statistics
8.3, 8.4
Project prelim info
Thu Feb 02 Sampling distribution of statistics
8.5
HW 3

5 Tue Feb 07 Sampling distribution of statistics
8.6
Project Outline

Thu Feb 09 Sampling distribution of statistics
8.7
HW 4

6 Tue Feb 14 Hypothesis testing
9.1, 9.5


Thu Feb 16 Hypothesis testing
9.6, 9.7
HW 5

7 Tue Feb 21 Hypothesis testing
9.7, 9.8


Thu Feb 23 Linear regression
11.1-11.3
HW 6

8 Tue Feb 28 Linear regression
11.4, 11.5
Project progress report
Thu Mar 02 Analysis of variance
11.6
HW 7
9 Tue Mar 07 Midterm Exam



Thu Mar 09 Frequentist model selection (goodness-of-fit)
10.1, 10.2 Project prelim slides due Friday

10 Tue Mar 14 Non-parametric methods
10.3, 10.4


Thu Mar 16 Final project presentations

Project final slides
Project final report due 3/17