Course Program:
Note: Since CmpE 544 was not offered last semester, CmpE 545 this semester will contain the most important topics of the two.
Course description: Neural networks are composed of interconnected processing units where the connection weights can be learned from data. We discuss various neural network architectures, learning algorithms, and applications.
Topics:
- Supervised Learning
- Bayesian Decision Theory
- Parametric Methods
- Nonparametric Methods
- Multilayer Perceptrons
- Deep Learning
- Radial Basis Functions and Mixture of Experts
- Combining Multiple Learners
- Reinforcement Learning
- Design and Analysis of Experiments
Textbook:
E. Alpaydın Introduction to Machine Learning, MIT Press, 3rd ed. 2014. You are not obliged to buy the book; lecture slides are available for download at https://www.cmpe.boun.edu.tr/~ethem/i2ml3e/
Grading:
- Homeworks: %30
- Project %40
- Final %30