|
Language of Instruction
|
English
|
|
Level of Course Unit
|
Bachelor's Degree
|
|
Department / Program
|
COMPUTER ENGINEERING
|
|
Type of Program
|
Formal Education
|
|
Type of Course Unit
|
Elective
|
|
Course Delivery Method
|
Face To Face
|
|
Objectives of the Course
|
Gain an understanding of pattern recognition methods
Learn the techniques used for developing pattern recognition models
Develop skills for practical aspects of deep learning
|
|
Course Content
|
This course provides an introduction to pattern recognition. It covers Bayesian and frequentist statistics, Bayesian learning methods, decision theory, generalized linear models and the exponential family, and regression models. Mathematical principles will be explained to provide a solid foundation for pattern recognition. Methods will be implemented by a software and applied on various pattern recognition problems.
|
|
Course Methods and Techniques
|
Grading Policy
The final grades will be computed based on the general performance of the class and the distribution of grades (i.e. who deserves A and who deserves F). The grading strategy will be a combination of the standard catalogue grading and curve grading.
Attendance Policy
Each student is expected to attend to at least 50% of the theoretical classes. If not he/she will get NA as the final grade.
Late Submission Policy
It is the student s responsibility to follow the classes and do the assignments on time. Late submissions will be subject to a penalty of 25% if submitted within one week after the due date and %50 if submitted after one week.
Make-Up Policy
There are no make-ups in homework assignments, labs and quizzes. The student may be exempt from these assignments if a written and formal documentation is provided. Possible reasons for excused absences include serious illnesses, illness or death of a family member, university related trips and other serious circumstances. Acceptable documents for claiming an excused absence include medical doctor’s statements, petitions related to official university travels, court related documents, etc. If the student misses an exam (midterms or final) he or she can take a make-up exam upon submitting a formal document.
|
|
Prerequisites and co-requisities
|
None
|
|
Course Coordinator
|
Prof.Dr. VEHBİ ÇAĞRI GÜNGÖR cagri.gungor@agu.edu.tr
|
|
Name of Lecturers
|
None
|
|
Assistants
|
None
|
|
Work Placement(s)
|
No
|
Recommended or Required Reading
|
Resources
|
Introduction, polynomial curve fitting, probability theory (LO1, LO2, LO3, LO4, LO5)
Model selection, curse of dimensionality (LO1, LO2, LO3, LO4, LO5)
Decision theory (LO1, LO2, LO3, LO4, LO5)
Information theory (LO1, LO2)
Probability distributions, binary variables, multinomial variables, Gaussian distribution (LO1, LO2)
Exponential family, nonparametric methods (LO1, LO2)
Linear models for regression, linear basis function models, bias-variance decomposition (LO1, LO2, LO3, LO4, LO5)
Bayesian linear regression (LO1, LO2)
Bayesian model comparison, evidence approximation (LO1, LO2)
Linear models for classification (LO1, LO2)
|
|
|
|
|
|
|
|
|
|
|
|
|
|