STAT 515: Theory and Applications of Nonlinear Models 

Class meets: MWF 9:00 am -- 9:50 am in: PHY 39

Instructor: Ranjan Maitra (Ron-joan Moi-tro)

Course Prerequisites:
  1. Stat 511
  2. Stat 543 or Stat 447
Grading Scheme: Course Description: This course is designed for Ph. D.-level and advanced Masters-level students. In this course, we will study extensions of linear models and into nonlinear models. We shall do this by developing theory and methodology of generalized linear models. We will extend our development to cover additive models and generalized additive models. Finally, we will cover the completely non-parametric regression and classification tree methodology. We will also apply our understanding of linear and nonlinear models into analyzing datasets derived from real-world applications. We will also study how to communicate our results and conclusions to non-statistical audiences.   Textbook: Because all the material is spread out over three to four books, there are no required textbooks for this class. However, the following books are recommended: Statistical Software: The statistical software used throughout this class will be R. R is very similar to Splus but comes under the GNU Public License. It is a comprehensive statistical software package freely available from R is developed by a team of international researchers and operates under the GNU Public License and is free. It is very similar, though not the exact same software as the commercially available Splus. Most commands in Splus work with R. All lab machines running Windows and Linux have R installed. Since the software is freely available, you may download it from the above web site and use it on your home computer. You may use either the Windows version or the Unix/Linux version. Please note that your installation of R is at your own risk, though the department systems administrators can perhaps help. Also, you may use Splus, in lieu of R, though the latter provides for more flexibility. Please note that SAS is often not an option for many of the topics we shall be covering. Homeworks: Homeworks will be handed out every two weeks. This will mostly consist of applying and exploring the concepts learnt in class. A considerable part of the homework will involve computer work. Examinations: There will be one midterm examination whose date wil be announced in consultation with the students. The scores in this exam will contribute to 20% of your final grade. The exam will have an in-class theory component and a take-home data analysis component. There will be a take-home final examination, worth 20% of the final grade and on the lines of a project, due on the last day of classes. Each student will have an oral defense of his/her project report on Wednesday, December 17, 2003 between 7:30 and 9:30 am (exact schedules to be announced later). Projects: There will be three data analysis projects assigned during this semester. The goal in each of these projects will be to learn how to analyze a given real-world data set, to understand the scientific or research question being asked of the statistician, to formulate and provide a reasonable statistical approach to the problem, and to present a well-written report to a non-statistical audience. The three projects are structured as follows: In the first project, we will have extensive discussions, led by the instructor, on how to proceed with the given dataset. For the second project, I will adopt a slightly more hands-free approach, where the students will discuss a reasonable approaches for the project. I will moderate the discussion as well as answer questions. Each of the first two projects is worth 15% of the final grade. The third and final project, together with the oral defense, is worth 20% of your grade. In this final project, you will be prohibited from discussing the matter with anyone excluding the instructor who may choose not to answer your questions. This final project will be the same as your final exam. Please note that for all three projects, you are individually responsible for performing the statistical analysis, and for writing the final report . Each report will be graded on a 15-point scale, with 5 points each for (a) the validity of the statistical analysis, (b) the scientific component, and (c) quality of the write-up in communicating the results to a non-statistical audience. Course Homepage: The course homepage will be located on the WWW at I will try and keep this homepage as upto date as possible. However, you are still responsible for any announcements made in class.