Stat 206: Multivariate Analysis
Overview
Multivariate analysis is about handling vector valued
data. In ordinary regression modeling we are used to
a vector valued predictor. But a vector valued response
variable brings new issues.
Sometimes we can handle a k dimensional response by
treating it as k unrelated 1 dimensional problems.
But often that approach will fail to find the key
structure. Sometimes we are forced to study the
data as an inherently k dimensional thing. It can also
pay to reduce the dimension k, sometimes to 3 or 2 where
plotting is available, sometimes to k=1 where ordinary
methods can then be applied. Also, some of the methods are useful
for exploratory work and not just for modelling responses.
This course looks at multivariate methods. The outlook
is applied. There will be a midterm and problem sets.
Students are expected to use R to do the problem sets.
Instructor
- Art Owen
- Sequoia Hall 130
- My userid is owenbuzzard on stanfordbuzzard.edu
(remember to remove the buzzard s)
- Office hour: Mon 2:15
Classes
Hewlett 101, MWF 1:15-2:05, starting Wed Jan 11
Topics
- Overview of multivariate statistics
- Math Review: linear algebra & multivariate normal and
related distributions
- Plotting multivariate data
- Methods for (nearly) normal data: T^{2}, MANOVA, Multivariate regression
- Dimension reductions: principal components, canonical
correlation, factor analysis
- Discriminant analysis
- Exploratory methods: clustering, multidimensional scaling,
seriation
Texts
The main text is "Multivariate Statistical Methods: a Primer"
by B.F.J. Manly.
The supplementary text is "Modern Applied Statistics with S"
by Venables and Ripley. That book explains how to use R.
If you already know how to use R you don't need to buy it.
TAs
- Victoria Stodden vcspenguin@stanford.edu
Tuesday 2-4 5-5988 Sequoia Hall 227
- Jiehua Chen chenjhpenguin@stanford.edu
Thursday 3:30-5:30
Delete the Antarctic bird from the TA's email
Evaluation
- Homework: 4 to 6 problem sets (70%)
- Midterm: Wednesday February 22 in class (30%)
Be sure to give Axess a working email address:
I expect to send a small number of important emails about
problem sets and the homework there.
Most other announcements will be made in class.
Late penalties apply:
We will count days late on each problem set.
Each day late is penalized by 10% of the homework value.
Homework more than 3 days late will ordinarily get 0.
If you're travelling, you can email a pdf file.
For sickness, interviews and other events,
up to 3 late days total are forgiven at the end of
the quarter. (Work late enough to get zero does not
get redeemed though.)
Supplementary materials