Stat 351 Nonparametric Statistics
Instructor: Moo K. Chung
Lectures: TR 2:30-3:45 1289 CS&S
Office Hour: TR 1:30-2:20 or by appointment
Office: 4382 CS&S Tel:
Requirements: Stat 201,301 or 224. Assignment problems
will require computer programming in matlab, S+/R or any other programming
Textbook: J.D. Gibbons and S. Chakraborti,
Nonparametric Statistical Inference, 3rd Edition, Marcel Dekker,
Inc., 1992. One copy has been asked to be reserved in the engineering library.
Basics: hypothesis testing, p-values (lecture
Order statistics: probability integral transform,
random number generator, empirical distributions, histograms (lecture 3,4).
Run test: the total number of runs, the length
of the longest run (lecture 5,6).
Goodness-of-fit test: Chi-squre, Kolmogorov-Smirnov
test (lecture 7-9)
Testing Normality: normal probability plot,
qq-plots (lecture 10)
Rank test: ranks, correlations (lecture 11)
Sign test: ordinary sign test, Wilcoxon signed-rank
test (lecture 12,13)
Two sample test:Kolmogorv-Smirnov test,
U test (lecture 14, 15)
Kruskal-Wallis H test (lecture 16)
Kernel methods: kernel density estimation,
smoothing, numerical implementation in one dimension (lecture 17-18)
Splines: Bezier curves, cubic splines, cublic
spline smoothing (lecture 19-20)
Basis function methods (lecture 21-23)
Final Grade (100%) = Assignments (30%) + Midterm
(20%) + Takehome Exam (20%) + Final Exam or Project (30%).
Assignments (30%) No late homework will be
Midterm Exam (40%) There will be one in class
midterms on October 15 or 17 and a takehome exam afterward. Calculators
are permitted and one page cheating sheet will be permitted. The
day to drop the course is November 1.
Final Exam or Project (30%) Covers everything.
Calculators and one page cheating sheet will be permitted.
projects in matlab or S+/R can substitute the final exam. Project topics
will be announced later. Students are required to submit at least 7 page
double spaced and typed report by the last class day ( December 12).
Computer Access and Softwares
Matlab code handouts
Creating an account in stat computers http://www.stat.wisc.edu/computing/instruct.html
Homeworks & Exams
statistics - probability integral transform
test - total run
test - stock price simulation
goodness of fit test
sample Kolmogorov-Smirnov test
kernel smoothing and density estimation