Syllabi Header

Math 763: Generalized Linear Models
Fall 2018 Graduate Course Syllabus

NJIT Academic Integrity Code: All Students should be aware that the Department of Mathematical Sciences takes the University Code on Academic Integrity at NJIT very seriously and enforces it strictly. This means that there must not be any forms of plagiarism, i.e., copying of homework, class projects, or lab assignments, or any form of cheating in quizzes and exams. Under the University Code on Academic Integrity, students are obligated to report any such activities to the Instructor.

Course Information

Course Description: Theoretical and applied aspects of generalized linear models. Classical linear models, nonlinear regression models, and generalized estimating equations.

Number of Credits: 3

Prerequisites: MATH 662 and MATH 665 or departmental approval.

Course-Section and Instructors

Course-Section Instructor
Math 763-001 Professor S. Dhar

Office Hours for All Math Instructors: Fall 2018 Office Hours and Emails

Required Textbooks:

Title Generalized Linear Models: With Applications in Engineering and the Sciences
Author Myers, et al.
Edition 2nd
Publisher John Wiley & Sons, Inc.
ISBN # 978-0470454633
Website NJIT Moodle

University-wide Withdrawal Date:The last day to withdraw with a w is Monday, November 12, 2018. It will be strictly enforced.

Course Goals

Course Objectives: This course teaches theory and practice of generalized linear models (GLM), testing, estimation, and confidence intervals of parameters, regression and analysis of variance, modeling nonlinear regression diagnostics and their plots, variable selection and model selection. Transformation method such as Box-Cox is taught. In case over dispersion is present, how to correct for it is introduced. Also when, data are dependent generalized estimating equations are introduced. Statistical software such as SAS and R, are extensively used to analyze data.

Course Outcomes: Upon successful completion of this course, the student will be able to:

Policies

DMS Course Policies: All DMS students must familiarize themselves with, and adhere to, the Department of Mathematical Sciences Course Policies, in addition to official university-wide policies. DMS takes these policies very seriously and enforces them strictly.

Grading Policy: The final grade in this course will be determined as follows:

Hand-in Homework and Quizzes 20%
Midterm Exam 25%
Course Project 25%
Final Exam 30%
Total 100%

Your final letter grade will be based on the following tentative curve.

A 90 - 100 C+ 75 - 79
B+ 85 - 89 C 65 - 74
B 80 - 84 F 0 - 64

Attendance Policy: Attendance at all classes will be recorded and is mandatory. Please make sure you read and fully understand the Math Department’s Attendance Policy. This policy will be strictly enforced.

Course Policy: It is required that the student read the textbook for the material already covered in class by the instructor and confirm that the basic solved problems are understood and practice solving textbook problems. More explicitly, students must work on the examples and exercises and problems from the textbook on the topics already covered in class, and learn to solve them correctly. The student should compare his or her answers with those given at the end of the textbook or by the instructor. Instructor holds the right to modify in class exams, homework, quizzes dates in the best interest of the class. Official announcements are made using NJIT student emails or emails provided by students to NJIT as official emails. Only basic calculators without graphic capabilities are allowed during exams and quizzes.

Homework Policy: Homework will be assigned roughly every week. Late homework will not be accepted.

Class Policy:

Exams: There will be one midterm and one final exam (see course outline). All exams will be closed-book. No make-up exams are allowed in case of extenuating circumstances with legitimate and verifiable excuse, scores will be imputed. Calculators are allowed but should be basic and without graphing capabilities.

Midterm Exam October 24, 2018
Final Exam Period December 15 - 21, 2018

The final exam will test your knowledge of all the course material taught in the entire course. Make sure you read and fully understand the Math Department's Examination Policy. This policy will be strictly enforced.

Makeup Exam Policy: To properly report your absence from a midterm or final exam, please review and follow the required steps under the DMS Examination Policy found here:

Cellular Phones: All cellular phones and other electronic devices must be switched off during all class times.

Additional Resources

Accommodation of Disabilities: Disability Support Services (DSS) offers long term and temporary accommodations for undergraduate, graduate and visiting students at NJIT.

If you are in need of accommodations due to a disability please contact Chantonette Lyles, Associate Director of Disability Support Services at 973-596-5417 or via email at lyles@njit.edu. The office is located in Fenster Hall, Room 260.  A Letter of Accommodation Eligibility from the Disability Support Services office authorizing your accommodations will be required.

For further information regarding self identification, the submission of medical documentation and additional support services provided please visit the Disability Support Services (DSS) website at:

Important Dates (See: Fall 2018 Academic Calendar, Registrar)

Date Day Event
September 4, 2018 T First Day of Classes
September 10, 2018 M Last Day to Add/Drop Classes
November 12, 2018 M Last Day to Withdraw
November 20, 2018 T Thursday Classes Meet
November 21, 2018 W Friday Classes Meet
November 22 - 25, 2018 R - Su Thanksgiving Recess
December 12, 2018 W Last Day of Classes
December 13 & 14, 2018 R & F Reading Days
December 15 - 21, 2018 Sa - F Final Exam Period

Course Outline

Week Sections Topic Assignment
1
(9/5, 9/10)
Chapter 1 and  2.1, 2.2.1 Linear regression (matrix formulation, ordinary least squares (OLS) estimator, Gauss Markov theorem.) Read Chapter 1. Assignment due week 2
2
(9/12, 9/17)
2.2.2-2.2.5 Linear regression models (other properties of the OLS estimator, estimation and hypothesis testing) Assignment due week 3
3
(9/19, 9/24)
2.2.6-2.5 Linear regression models (residual diagnostics, maximum likelihood estimation [MLE], generalized least squares) Using R and SAS to perform Regression Analysis. Assignment due week 4
4
(9/26, 10/1)
2.6-2.7 Linear and nonlinear regression models (weighted least squares, estimation in nonlinear regression models) Assignment due week 5
5
(10/3, 10/8)
3.1-3.7 Nonlinear regression models (Gauss-Newton method, inference, weighted nonlinear regression) Assignment due week 6
6
(10/10, 10/15)
4.1-4.2.6 Logistic regression models (model description, MLE and dispersion properties, likelihood ratio inference) Assignment due week 7
7
(10/17, 10/22)
4.2.7-4.4 Logistic and Poisson regression models (odds ratios, estimation and inference for Poisson regression) Assignment due week 9
8 MIDTERM EXAM October 24, 2018
9
(10/29, 10/31)
5.1-5.4 GLM (components of a GLM, exponential family of distributions, formal structure for the class of GLMs, likelihood equations for GLMs, an algorithm for fitting GLMs, quasi-likelihood) Assignment due week 10
10
(11/5, 11/7)
5.5-5.7 GLM (the gamma family, canonical and log links for the gamma family, a class of link functions -- the power function, inference for GLMs) Assignment due week 11
11
(11/12, 11/14)
5.8-5.9 Examples with gamma family, Using R Assignment due week 12
12
(11/19, 11/26)
5.10-5.11 GLM and data transformation, Modeling Processes Mean and Variance Assignment due week 13
13
(11/29, 12/3)
5.12 Quality of asymptotic results. STUDENTS' PROJECT PRESENTATIONS Assignment due week 14
14
(12/5, 12/10)
6.1-6.3 Generalized estimating equations (residual analysis for GLMs, layout for longitudinal studies, correlation matrix, identity link, examples) STUDENTS' PROJECT PRESENTATIONS Assignment due week 15
15
(12/12)
Review
Final Exam (12/15 to 12/21)

Updated by Professor S. Dhar - 8/31/2018
Department of Mathematical Sciences Course Syllabus, Fall 2018