COURSE UNIT TITLE

: SPECIAL TOPICS IN UNSUPERVISED ADAPTIVE FILTERING

Description of Individual Course Units

Course Unit Code Course Unit Title Type Of Course D U L ECTS
EEE 5119 SPECIAL TOPICS IN UNSUPERVISED ADAPTIVE FILTERING ELECTIVE 3 0 0 7

Offered By

Graduate School of Natural and Applied Sciences

Level of Course Unit

Second Cycle Programmes (Master's Degree)

Course Coordinator

ASSOCIATE PROFESSOR MEHMET EMRE ÇEK

Offered to

ELECTRICAL AND ELECTRONICS ENGINEERING NON -THESIS (EVENING PROGRAM)
ELECTRICAL AND ELECTRONICS ENGINEERING
ELECTRICAL AND ELECTRONICS ENGINEERING
ELECTRICAL AND ELECTRONICS ENGINEERING

Course Objective

The main purpose of this lesson is to provide the fundamentals of the blind adaptation methods where the desired response is not needed to perform filtering process differing from the linear (supervised) adaptive filtering methods.

Learning Outcomes of the Course Unit

1   The students are expected to describe the source separation problem and basics separator structures used for unsupervised filtering problem.
2   The students are expected to formulate the convolutional noisy mixtures, and to perceive the pre-processing techniques such as whitening, PCA and Noise Reduction.
3   The students are expected to understand the theory of Independent Component Analysis (ICA) and to realize the ICA algorithm to separate the mixed unknown sources.
4   The students are expected to establish the relation between blind source separation and blind deconvolution techniques.
5   The students are expected to computationally solve blind source separation and blind deconvolution problems used in signal and image processing applications

Mode of Delivery

Face -to- Face

Prerequisites and Co-requisites

None

Recomended Optional Programme Components

None

Course Contents

Week Subject Description
1 Why Adaptive Filtering Supervised and Unsupervised Forms of Adaptive Filtering
2 Standart Gradient Adaptotion, Natural Gradient Adaptation.
3 Whitening Process, Principal Component analysis, Eigendecomposition, Singular Value Decomposition, Noise Reduction.
4 Blind Extraction of Source Signals.
5 Entropic Constrasts for Sour Separation Geometry and Stability, Constrast Functions for Source Separation.
6 Introduction to Independent Component Analysis (ICA). Theory of ICA, Basic Concepts.
7 Maximizing Non-Gaussianity, Fast ICA Method, Applications of ICA
8 Blind Separation of Delayed and Convolved Sources
9 Blind Deconvolution of Multipath Mixtures
10 Relationships Between Blind Deconvolution and Blind Source Separation.
11 Blind Separation of Independent Sources Based on Multiuser Kurtosis Optimization Criteria.
12 Discussions on Term Projects
13 Discussions on Term Projects
14 Term Project Presentations

Recomended or Required Reading

1) S. Haykin S. Unsupervised Adaptive Filtering Blind Source Separation, Wiley Interscience Publication, 2000,
2) S. Haykin S. Unsupervised Adaptive Filtering Blind Deconvolution, Wiley Interscience Publication, 2000,

Planned Learning Activities and Teaching Methods

Lectures, presentations

Assessment Methods

SORTING NUMBER SHORT CODE LONG CODE FORMULA
1 ASG ASSIGNMENT
2 PRS PRESENTATION
3 FCG FINAL COURSE GRADE ASG * 0.50 + PRS * 0.50


*** Resit Exam is Not Administered in Institutions Where Resit is not Applicable.

Further Notes About Assessment Methods

Homeworks, Term Project Report

Assessment Criteria

To be announced.

Language of Instruction

English

Course Policies and Rules

To be announced.

Contact Details for the Lecturer(s)

emre.cek@deu.edu.tr

Office Hours

To be announced.

Work Placement(s)

None

Workload Calculation

Activities Number Time (hours) Total Work Load (hours)
Lectures 14 3 42
Preparations before/after weekly lectures 12 6 72
Preparing assignments 6 6 36
Preparing Term Project 1 25 25
TOTAL WORKLOAD (hours) 175

Contribution of Learning Outcomes to Programme Outcomes

PO/LOPO.1PO.2PO.3PO.4PO.5PO.6PO.7PO.8PO.9PO.10PO.11PO.12PO.13PO.14
LO.144
LO.2553
LO.3553
LO.4553
LO.545