analysis. Topics will include statistical graphics, advanced regression techniques, curve fitting

and smoothing, generalized additive models, CART, multivariate techniques, cross-validation

and the bootstrap. Additional topics may include survival analysis, simulation methods and

neural networks. Prerequisite: MATH 165 and STAT 360.

Modern Applied Statistics in S-Plus (3^rd ed.) – Venables and Ripley, Springer-Verlag, 2000.

Prepared handouts.

The purpose of the Oral Communication Flag requirement is to complete the process of providing graduates of Winona State University with the knowledge and experience required to enable them to become highly competent communicators by the time they graduate. Courses can merit the Oral Communication Flag by demonstrating that they allow for clear guidance, criteria, and feedback for the speaking assignments; that the course requires a significant amount of speaking; that speaking assignments comprise a significant portion of the final course grade; and that students will have opportunities to obtain student and faculty critiques of their speaking.

1. earn significant course credit through extemporaneous oral presentations;

The final project for this course is to present orally the results of large data analysis the students conduct. The format for this talk will be similar to those used at professional meetings in the field of statistics. Along with a written report this project constitutes 50% of the students final grade.

2. understand the features and types of speaking in their disciplines;

Students will learn the essential features of talk where statistical analysis is discussed..

3. adapt their speaking to field-specific audiences;

Clearly their talks would be directed at a statistically mature audience. In addition, the data they will be asked to analyze will come from a different field of study thus students will have to discuss the findings in the context of a second discipline.

4. receive appropriate feedback from teachers and peers, including suggestions for improvement;

Student presentations will be made in front of their classmates and when feasible other faculty will be invited to attend. In some cases their presentations may be in the form of a departmental seminar.

5. make use of the technologies used for research and speaking in the fields;

Students will definitely be using a computer with projection capabilities in making their presentations. This way students can interactively demonstrate aspects of their statistical analysis. Also students will have the opportunity to use presentation software such as Powerpoint.

6. learn the conventions of evidence, format, usage, and documentation in their fields.

The oral presentation, final written report and other class projects promote the students abilities in all of these areas.

1. Scatterplots and Scatterplot Matrices

2. Spin Plots

3. Boxplots

4. Coplots and other Trellis Displays

5. Image plots

6. 3-D Surface plots

C. Regression Techniques

1. Regression models in S-Plus

2. Interpretation of results

3. Checking model assumptions

4. Violations and possible corrections

5. Model Selection strategies

D. Modern Regression Methods and Related Topics

1. Introduction and motivation

2. Kernel smoothing

3. Spline smoothing

4. Supersmoother

5. Projection pursuit

6. ACE, AVAS and others

7. Model validation via cross-validation

8. Use and interpretation of results

E. Local Regression Models

F. Generalized Additive Models

G. Classification and Regression Trees

1. Introduction and motivation

2. Regression tree methodology

3. Graphical displays and interaction

4. Interpretation

H. Multivariate Methods

1. Advanced clustering Procedures

2. Flexible discriminant methods

3. Visualizing higher dimesional data

I. The Bootstrap

1. Estimation

2. Hypothesis Testing

Basic Instructional Plan and Methods Utilized : The basic method of instruction will be lecture, discussion and laboratory work.

Course Requirements: Course requirements will include a number of data analysis projects. The final course project is to perform a large data analysis using the some of the methods discussed. Students will present their results of this analysis both orally and in written report.