DSC 704 - Nonlinear Models for Data Science

Semester/Academic year
Contact Information
  • Instructor:
  • Office:
  • E-Mail:
  • Phone:
  • Office Hours:
Course Description

This course provides an understanding of the application of software technologies that enables users to make better and faster decisions based on nonlinearity in big data features. This course covers the statistical tools needed to understand empirical research and to plan and execute independent research projects. Topics include statistical inference in nonlinear models, nonlinearity in regression, dimension reduction, data clustering, similarity, neighbours and homogeneity/heterogeneity and evaluation of policies and programs. Students will learn the principles and best practices for how to use big data in order to support fact-based decision-making. Emphasis will be given to applications in various data which has nonlinearity in big data.

Course Outcomes

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

  1. define and model nonlinearity in the data structure under investigation; DSLO 1
  2. use mathematical models and solve for equilibrium in nonlinear models. Also models will be used analyze the policies related to various research field. - DSLO: 2,3,4
  3. analyze and critically evaluate from oral written, and visual materials.- DSLO: 3,6,8
  4. have the ability to predict the effects of changes in any kind of policy related to investigated field. –DSLO: 3,7,9,11

The course outcomes are assessed by using a case study, quizzes, two mid-term exams and a comprehensive final exam.

DS Learning Outcomes:

Upon successful completion of the program, students will be able to:

  1. be competent in using appropriate nonlinearity in big data modelling.
  2. define, formulate and analyze a complex nonlinearity in data structure involving human, material, machinery, money, information, time energy elements and various others and design it under realistic constraints and conditions.
  3. design and conduct experiments, gather data, analyze and interpret results for investigating complex nonlinearity in data structure research questions.
  4. be proficient in statistical analysis of nonlinear data and data management and apply data science concepts and methods to solve problems in various field.
  5. use information technologies effectively with the knowledge of state-of-the art hardware, and software capabilities related to nonlinearity in big data.
  6. communicate effectively, using information technology and oral and written skills to enhance decision making process through better communication.
  7. make ethical and legal decisions by considering cultural differences.
  8. work efficiently in interdisciplinary and multidisciplinary teams by collaborating effectively, in addition to an individual effective working ability.
  9. enhance critical thinking skill by integrating relevant information, decision-making techniques, and concepts through the interdisciplinary nonlinearity in big data science area.
  10. recognize the importance of data modelling for entrepreneurship, innovation sustainable development and various other fields.
  11. have knowledge of the global and social effects of nonlinearity in data science and proper modelling of the nonlinearity of the data in various field.

The core courses, DS 705, have to be passed in order to take this course as well as for taking the other courses.

Required Text(s) and Materials
  1. W.W.S. Wei (1991) Time Series Analysis: Univariate and Multivariate Methods. Addison Wesley Publishing Company.
  2. J. D. Hamilton (1994)Time Series Analysis. Princeton University Press Terasvirta T. and Granger C.E.J Modelling Nonlinear Economic Time Series (Advanced Texts in Econometrics) 1st Edition.(1995)
  3. Smith R.P. and Fuertes A.M. Panel Time Series (2012)
  4. Breitung, J. & Pesaran, M.H., 2005. "Unit Roots and Cointegration in Panels," Cambridge Working Papers in Economics 0535, Faculty of Economics, University of Cambridge.
  5. Uçar N. and Omay T.,(2009) “Testing For Unit Root In Nonlinear Heterogeneous Panels” Economics Letters. 104(1), p. 5-7.
  6. Omay, T. and Kan E. O., (2010) “Re-examining the Threshold Effects in the Inflation-Growth Nexus: OECD Evidence” Economic Modelling. 27(5), p. 996-1005
  7. Emirmahmutoglu, F. Omay, T., (2014) "Reexamining the PPP hypothesis: A nonlinear asymmetric heterogeneous panel unit root test", Economic Modelling, 40(C), p. 184-190.
  8. Omay, T. Hasanov, M. and Ucar, N., (2014) “Energy Consumption and Economic Growth: Evidence From Nonlinear Panel Cointegration and Causality Tests” Applied Econometrics, vol. 34(2), p. 36-55 (Russian SSCI)
  9. Omay, T., Yuksel, A., Yuksel, A., (2015) "An empirical examination of the generalized Fisher effect using cross-sectional correlation robust tests for panel cointegration," Journal of International Financial Markets, Institutions and Money, Elsevier, 35(C), p. 18-29.
  10. Omay, T., Apergis, N., Özçelebi, H., (2015) "Energy Consumption And Growth: New Evidence From A Non-Linear Panel And A Sample Of Developing Countries", The Singapore Economic Review, 60(02), p. 1-30.
  11. Omay, T. (2015) “Fractional Frequency Flexible Fourier Form to approximate smooth breaks in unit root testing”, Economics Letters, 134(C), p. 123-126.
  12. Omay, T., Yuksel, A., Yuksel, A.,(2016) “A Note on the examination of the Fisher hypothesis by using panel co-integration tests with break” Journal of Economic Forecasting. 19(2), 13-26.
  13. Çorakcı A., Emirmahmutoğlu, F., Omay, T., (2017) “Re-examing the real interest rate parity hypothesis (RIPH) using panel unit root tests with asymmetry and cross-section dependence” Journal of European Economics: Empirica, 44(1), p. 91 – 120.
  14. Omay, T., Çorakcı A., Emirmahmutoğlu, F., (2017) “Real interest rates: nonlinearity and structural breaks” Empirical Economics, 52(1), p. 283-307.
  15. Omay, T., Emirmahmutoğlu, F., (2017) “The comparison of power and optimization algorithms on unit root testing with smooth transition”, Computational Economics, 49(4), p. 623 – 651.
  16. Omay, T., Emirmahmutoğlu F., Z. S. Denaux, (2017) Nonlinear Error Correction Based Cointegration Test in Panel Data, Economics Letters.
  17. Omay, T., Eyden, R., Gupta, R.(2018) "Inflation-Growth Nexus in Africa: Evidence from a Pooled CCE Multiple Regime Panel Smooth Transition Model," Empirical Economics.
  18. Omay, T., Hasanov, M., Shin, Y.,(2018) “Testing for Unit Roots in Dynamic Panels with Smooth Breaks and Cross-sectionally Dependent Errors” Computational Economics.
  19. Omay, T. (2014) “A Survey about Smooth Transition Panel Data Analysis” Econometrics Letters, 1(1), p. 18-29.
Assessment Method(s) and Evaluation

Grading will be based upon 10 short exams, 1 mid-term exam, 1 term-project and 1 final exam.

Mid-term Exam:

This is 40% of the final average: Two mid-term exams will be administered by means of the portal on the scheduled date. Each exam will be worth equal point (20% each). Each mid-term exam consists of multiple choice questions and short work problem. There will be 1.5 hours to complete it.

Final Exam:

This is 35% of the final average: A comprehensive final exam will be administered by means of the portal on the scheduled date. The comprehensive final exam consists of multiple choice questions and short work problem. You only have a two-hour to complete it.


Both midterm exams and final exam also have one bonus question for extra credit. Questions in the exams are designed to make sure that you understand the basic economics tools used in air transportation. The types of questions on the exams will be similar to those asked in the study questions and the class materials covered during the class period.Since the final exam is cumulative, the materials covered after the second midterm exam will be given more weight in the final exam.


This is 10% of the final average: Students are required to complete 10 online quizzes throughout the course by means of the portal. The online quizzes are composed of ten multiple-choice questions with each quiz covering a separate chapter. The overall score becomes available to each student upon completion of his or her test. Quizzes can be done ahead of time, but it cannot be made up after the deadline of each particular quiz. There will be no make-up online quiz if you fail to complete it by the deadline.

Case Study:

This is 15% of the final average: These assignment is aimed to improve the understating of current issues. Every student will be responsible for handing a topic in Big Data Analysis . The case studies will be assigned to the student by the second week of semester or the students have the option to determine their own topic which must be approved by the instructor. On an assigned day of the week, assigned student will upload the case study to the portal as a write-up. The write-up of the case study is due on the last day of class. Grades will be based upon instructor evaluation of your write-up. More details will be given during the course of the semester.

Policy on Make- ups: Make-up examinations will only be administered to students with excused absences. Excused absences include death in the immediate family, University sponsored trips or critical illness. Verification is required and permission to miss an examination must be secured PRIOR TO the scheduled examination time. If this condition is not met, a zero will be given for the missed exam.

Grading Scale
Grade Quality Points
A = Excellent 90 – 100%
B = Good 80 – 89%
C = Satisfactory 70 – 79%
D = Passing 60 – 69%
F = Failing below 60%
Incompletes- I

Incompletes (I) demonstrate that a student was doing sufficient work at the end of a semester period but, for reasons beyond the control of the student, was unable to complete all requirements of the course in the related semester. The grade I obliges student to complete all course requirements within a time period that is specified by the instructor. This time period can’t exceed one academic calendar year from the end of the semester in that the grade I is assigned. The students has to arrange with the course instructor in order to complete all course requirements in a specified time period. If all course requirements are not completed by the students in a specified time period, I is changed to the grade F, unless the instructor has assigned a different grade.


Students may withdraw from courses following the drop and add period until mid-term by completing the withdrawal process on the portal. A grade of "W" will appear in the student's official records if the student has withdrawn according to the SFU’s Withdrawal Policy. (Please see the SFU’s Withdrawal Policy for details.)

Attendance Policy

Participation and consistent attendance is essential for acceptable performance in the course. The students are expected to be present each class period except when special hardships occur, e.g. death in the immediate family, University sponsored trips or critical illness.
Regulations for attendance of Suje Florida University will be applied for this class.

Academic Integrity

Academic integrity is the responsibility of all Suje Florida University faculty and students. Cheating and plagiarism are not tolerated and will result in a failing grade, if the student is found guilty of cheating. All students are expected to do the work on their own and to accept standards of academic ethics.

Course Expectations
  1. As a portal course, it requires extensive work be done by students using the Internet. You must familiarize yourself with your portal account. Supplemental materials, including lecture notes will be posted on portal.
  2. Students are expected to read assigned material(s) prior to lecture and participate in discussions and activities.
  3. Log on at least three times a week – on different days in order to completely weekly assignments, assessments, discussions and/or other weekly deliverables as directed by the instructor.
  4. Check your e-mail often.
  5. Communications with the instructor should be via portal or e-mail. Email is preferred.
  6. It is your responsibility to ensure you are registered throughout the course.
  7. Discussion must always be civil. Rudeness or disrespect of other students will not be tolerated. We will respect the thoughts and opinions and others, even when we do not agree or believe the other person is terribly misinformed.
  8. Changes may be necessary in the syllabus. Students will be informed of changes to the syllabus.
  9. Students are responsible for any new material or announcements missed due to the absence.
  10. Students will use e-mail to communicate with the instructor. Emails sent Monday through Friday will be answered within 24 hours. Emails sent Saturday– Sunday may not be answered until Monday. If your email is not written in a respectful manner, you should not expect a response.
Tentative Detailed Course Content and Recommended Readings
Week Topic Recommended Reading(s)

Data Structure and types in nonlinear models



Nonlinear Static Models

Article 4-20


Nonlinear Dynamic Models

Article 4-20


Nonlinear Panel Unit roots First generation



Nonlinear Panel Cointegration First generation

  • - Article 4-20
  • - SRP and AM Chp3
  • - B. J. and P M.H

Cross Section Dependence: Factor Models and Nonlinearity

  • - Article 4-20
  • - SRP and AM Chp4

Cross Section Dependence: Estimators and Nonlinearity

  • - Article 4-20
  • - SRP and AM Chp5

Nonlinear Panel Unit roots Second generation

  • - Article 4-20
  • - SRP and AM Chp3
  • - B.J. and P M.H

Nonlinear Panel Cointegration Second generation

  • - SRP and AM Chp3
  • - B.J. and P M.H

Dimension Reduction with Nonlinear Panel Models

  • - Article 4-20
  • - Lecture Notes Available
  • - Different Forms of Nonlinearity
  • - Threshold Models
  • - Article 4-20
  • - Lecture Notes Available

Smooth Transition Models

  • - Article 4-20
  • - Lecture Notes Available
  • - Markov Switching Models
  • - Fourier Transforms
  • - Article 4-20
  • - JDH
  • - WWSW
  • - Hybrid Modelling of Nonlinearity
  • - Time varying and state dependent nonlinearity I
  • - Article 4-20
  • - Lecture Notes Available
  • - Hybrid Modelling of Nonlinearity
  • - Time varying and state dependent nonlinearity II
  • - Article 4-20
  • - Lecture Notes Available


Note: WWSW indicate W.W.S. Wei (1991) Time Series Analysis: Univariate and Multivariate Methods. Addison Wesley Publishing Company.

  • JDH indicate J. D. Hamilton (1994)Time Series Analysis. Princeton University Press Terasvirta T. and Granger C.E.J Modelling Nonlinear Economic Time Series (Advanced Texts in Econometrics) 1st Edition.(1995)
  • TT and GCEJ indicate Terasvirta T. and Granger C.E.J Modelling Nonlinear Economic Time Series
  • SRP and AM Smith R.P. and Fuertes A.M. Panel Time Series (2012)
  • B. J. and P M.H., Breitung. J. and Pesaran M.H., 2005. "Unit Roots and Cointegration in Panels," Cambridge Working Papers in Economics 0535, Faculty of
Student Opinion of Instruction

At the end of the term, all students will be expected to complete an online Student Opinion of Instruction survey that will be available on portal. Students will receive an email notification through their VSU email address when the SOI is available. SOI responses are anonymous to instructors/administrators. Instructors will be able to view only a summary of all responses two weeks after they have submitted final grade

Title IX Statement

Suje Florida University is committed to creating a diverse and inclusive work and learning environment free from discrimination and harassment. Discrimination on the basis of race, color, ethnicity, national origin, sex (including pregnancy status, sexual harassment and sexual violence), sexual orientation, gender identity, religion, age, national origin, disability, genetic information, or veteran status, in the Suje Florida University's programs and activities is prohibited as required by applicable laws and regulations such as Title IX. The individual designated with responsibility for coordination of compliance efforts and receipt of inquiries concerning nondiscrimination policies is the University's Title IX Coordinator.

Access Statement

Students with disabilities who are experiencing barriers in this course may contact the Access Office for assistance in determining and implementing reasonable accommodations.