Detailed_syllabus_03-22-10

Detailed_syllabus_03-22-10 - PUBH 8400 Sec 001 Richly...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
PUBH 8400 Sec 001, Richly Parameterized Models, Spring 2010 Detailed syllabus 3/22/10, Jim Hodges “RWC” is Ruppert, Wand, and Carroll’s book Semiparametric Regression . The book's web site is http://www.stat.tamu.edu/~carroll/semiregbook/ . This web site includes most of their datasets, an errata sheet (5 pages!), and computer code in SAS, WinBUGS, R/S+ and Matlab. Part I. Mixed Linear Models, per RWC Review: Read RWC Chapter 1, 2 (thru section 2.6) A. Mixed linear models in the standard formulation Reading: RWC Chapter 4 (except section 9), Chapter 16 1. The mixed linear model, standard form. Examples. 2. Doing statistics with this model a. Conventional (likelihood-based) analyses b. Bayesian analyses c. Pros and cons of the two approaches 3. A bit about computing Datasets: Pig weight data (RWC); molecular structure of a virus (Peterson et al 2001); rating vocal-fold images B. The constraint-case formulation; Lee & Nelder alternative syntax; measures of complexity Reading: Hodges (1998) sections 1, 2; Cui et al (2010). 1. The constraint-case formulation for richly-parameterized models 2. Lee, Nelder, and Pawitan’s book in 5 minutes 3. Measuring model complexity: degrees of freedom in the whole model fit; degrees of freedom for individual components of the model; residual degrees of freedom; prior distributions on degrees of freedom. C. Smoothing using penalized splines represented as mixed models Reading: RWC Chapter 3; Chapter 4 section 9; Chapter 6 sections 1-4. 1. Basics of smoothing splines: bases, knots, penalized splines, rank of smoothers. 1 3/22/10 PubH 8400/1 Richly Parameterized Models, Detailed Syllabus
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
2. Penalized splines as mixed linear models; fitted values, inference, complexity. Datasets: Global mean surface temperature. D. Additive and interaction models represented as mixed models Reading: RWC Chapters 7, 8, 9; Chapter 12 sections 1-3; Hodges et al (2007). 1. Additive models (RWC Chapter 7-9), starting with the simplest such model and building up to more complicated ones by adding spline smooths for effects, curves as random effects. 2. Interactions: categorical-by-continuous (RWC Chapter 12 sections 1-3); categorical-by- categorical (smoothed ANOVA; Hodges et al (2007)) Dataset: Physical properties of pig jawbone. E. Spatial smoothing using mixed linear models Reading: Review Cui et al (2010); RWC Chapter 13 sections 1-4 1. Smoothing on a lattice: conditional autoregressive (CAR) models/priors; using a lattice smoother to smooth a spatial effect in smoothed ANOVA. 2.
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 11/21/2011 for the course PUBH 8400 taught by Professor Jimhodges during the Spring '10 term at Minnesota.

Page1 / 6

Detailed_syllabus_03-22-10 - PUBH 8400 Sec 001 Richly...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online