Statistics 248 - Some Details - Spring 2012

Analysis of Time Series

Subtitle: Random Process Data Analysis: Models and Methods

Some specifics of the course and the project

Meant for graduate students in Statistics and other Departments.

Instructor: David Brillinger, brill@stat.berkeley.edu
Office hours: Wednesday 15:00 - 17:30 in 417 Evans

GSI: Jack Kamm
Office hours:Tues 3:30-5:30 in 307 Evans; Wed 1:00-3:00 in 387 Evans

Jan 17 class

Jan 19 class

Jan 24 class "handout"

Jan 26 class Measuring the association of point processes

Feb 2 class

Feb 7 and 9 classes

Feb 16 class

Feb 21 class

Feb 23 and 28 classes

Mar 1 and 6 classes

"Trend analysis: binary-valued and point cases"

"On Chinese earthquake history - an attempt to model an incomplete data set by point process analysis"

Mar 8 class

Mar 13 class

Mar 13 classs

Mar 15 classs

Mar 15 classs

Mar 20 classs

Mar 22 classs

Multiple regression notes

Nonlinear LS example B. A. Bolt and D. R. Brillinger (1979) Estimation of uncertainties in eigenspectral estimates from decaying geophysical series.

April 3 class

April 5 class

Remaining class materials

Partial syllabus

Some pertinent books

Some previous projects

Syllabus:
Random process data analysis: frequency- and time-side analysis of ordinary time series, stationarity and non-stationarity, parametric and nonparametric (ARIMA, GARCH, ...), Markov chains, point processes, spatial processes, spatial-temporal processes, ...
These topics will be presented in a comparative fashion

Classes: Tu Th 2:00 - 3:30 in 344 Evans

Lab Section: Friday 15:00 to 17:00 340 Evans

Lab website: TBA

Course Homepage: www.stat.berkeley.edu/~brill/Stat248

Supplimentary texts: D. R. Brillinger, Time Series: Data Analysis and Theory, SIAM.
P. Guttorp, Stochastic Modelling of Scientific Data, Chapman and Hall.

web partial version

Prerequisites: Statistics 101&102 or 134&135 or equivalents.

The course work will be directed towards the students preparing an analysis of pertinent scientific data using the methods covered in the course.

The Lab will be directed to teaching the pertintent material concerning the statistical package R, particularly the time series functions, and to helping the students with realizing their projects.

The grade will come from a combination of a Project Proposal and an Independent Project.

The Proposal is due March 15 at class and the Project is due by 14:10 May 3 in the class room or under my office door.

22 April 2012

brill@stat.Berkeley.EDU