Course: Machine Learning in Signal Processing

Code: 3ФЕИТ05010

ECTS points: 6 ECTS

Number of classes per week: 3+0+0+3

Lecturer: Assoc. Prof. Dr. Tomislav Kartalov

Course Goals (acquired competencies): The students that finish this course, should be able to: – Decompose, analyze, classify, detect and consolidate signals – Develop appropriate models for measured signals/data – Choose the appropriate tool for feature extraction – Assess / Evaluate the advantages and limitations of different signal processing tools for a given problem – Derive the supervised and unsupervised learning techniques studied in class – Choose an appropriate learning algorithm for a given problem – Develop basic supervised and unsupervised learning models – Assess / Evaluate the advantages and limitations of different machine learning algorithms.

Course Syllabus: Representing Sounds and Images. Introduction to Linear Algebra. Signal Representations – Component Analysis. Eigen representations: Eigenfaces. Boosting. PCA. ICA. NMF. Sparse NMF. Clustering. SVM. Mixture Models and EM. Linear Regression. Logistic Regression. Markov and Hidden Markov Models. Neural Networks. Deep Learning. Convolutional Networks.

Literature: