Information
Neptun ID, timetable, classroom
Description
Requirements
Attendance of the lectures is highly recommended. Attendance is registered. In-class performance is assessed and its results form part of the end-term grade. The final score (%) of the course is constructed as follows: 20% lecture and seminar attendance, 30% in-class test 50% homework/project presentation (last two lectures of the semester).
Grades (based on percentages)
- 80-100: 5 (excellent)
- 70-79: 4 (good)
- 60-69: 3 (medium)
- 50-59: 2 (satisfactory)
- 0-49: 1 (fail)
Online available material
see Coospace
Syllabus
-
Lecture 1
Motivational examples and introduction to optimization, examples, basic definitions. Lecture 1 slides
-
Lecture 2
Unconstrained optimization: ptimality conditions, convexity and optimization Lecture 2 slides
-
Lecture 3
Unconstrained optimization: the general line search method. Newton's method, Gradient descent method. Lecture 3 slides
-
Lecture 4
Quadratic programing: least squares method, lasso and ridge regression. Conjugated gradient method. Lecture 4 slides
-
Lecture 5
Constrained optimization I: Lagrange dual function, duality theorems, KKT theorem. Lecture 5 slides
-
Lecture 6
Constrained optimization II: penalty and barrier methods, applications. Lecture 6 slides
-
Lecture 7
Optimization for regresssion: Stochastic gradeint descent, AdaGrad, RMSprop. Lecture 7 slides
-
Lecture 8
Optimization for classification: logistic regresion, support vector machine. Lecture 8 slides
-
Lecture 9
introduction to global optimization: gradient based and direct search methods Lecture 9 slides
-
Lecture 10
Introduction to intervall arithmetic and reliable optimization Lecture 10 slides