STAT 579: Selected Topics: Spatial Statistics.

Department of Mathematics and Statistics, UNM
Spring Semester 2009

INSTRUCTOR

Gabriel Huerta
Office: 424 Humanities Building
email: ghuerta at stat dot unm dot edu
Phone: 277-2564
Class Time: T R 14:00-15:15.
Classroom: 424 Humanities
Course Web-Page: http://www.stat.unm.edu/~ghuerta/stat579/course.html
Office Hours: T R 15:30-17:00 or by appointment
*Please send me a note by e-mail or call me to set an appointment outside office hours.

DESCRIPTION

This main goal of this class is to learn about modern techniques in spatial statistics and spatio-temporal analysis using hierarchical methods. Topics that will be coverd are spatial models, variograms, stationarity, areal data models, MCMC and Bayesian hierarchical methods. Analysis will be performed using computer software such as R, GeoR, Winbugs and Sp-Bayes. The range of applications is general with perhaps emphasis on environmental data analysis and biostatistical applications.

PREREQUISITES

STAT 561 and STAT 540. Previous exposure to statistical computating and Bayesian methods will be helpful but no required. A related class is STAT 477/577 Introduction to Bayesian Modeling.

TEXTBOOK (required)

  • Banerjee, S, Carlin B. and Gelfand A. (2004) Hierarchical Modeling and Analysis for Spatial Data. Chapman and Hall/CRC.
    • The data sets of the book are available here
    Link to Dr. Brad Carlin's page
    GRADING

    The grading will be based on homework assignments, class discussion and presentations, a class project. Homeworks will be assigned approx. every other week and will involve some 'theory' and 'computer' exercises from the textbook. I expect homeworks to be presented in time (no late HW please) and as neatly as you can, including all relevant information: graphs, detailed proofs, discussion of results. The class project could be a spatial data of interest to you. It could also involve a research paper in "hierarchical spatial/spatio temporal modeling" The class presentation will be about your project and will take place during the last two weeks of the semester. A written version of your project will be due on Finals' week. Regular homework is worth about 50% of the course grade. The final project and class presentation is worth 50% of the course grade.

    HOMEWORK

    Will be posted here.

    SOFTWARE LINKS

    There will be a some amount of computation in this class. For that purpose we will use software as R, Bugs/Winbugs. Here are few links related to this item.

    R software
    Textbook's Data sets and code
    Bugs/Winbugs software
    GeoBUGS software
    The R library spBayes

    SCHEDULE

    There will be a certain amount of 'traditional' lectures but I also a lot of participation from you in the form of discussion of advanced sections of the book of other books or research papers, presentation in class of key HW problems or key issues with computer software. The lectures will focus on the following material

  • Overview of Spatial data problems. Ch. 1 from text.
  • Basics of point referenced data. Ch. 2.
  • Basics of area data model. Ch. 3.
  • Basics of Bayesian Inference. Ch. 4.
  • Hierarchical modeling for univariate spatial data. Ch. 5