EE Department Intranet -
Close window CTRL+W

ELEC70001 Adaptive Signal Processing and Machine Intelligence

Lecturer(s): Prof Danilo Mandic


Many areas in engineering, biomedicine and economy increasingly rely on learning algorithms which operate sequentially and in real time in order to solve problems that are not amenable to classic analyses. The aim of this module is to provide you with in-depth knowledge of the theoretical basis and applicability of modern methods for spectral estimation, algorithms which underlie adaptive signal processing, and machine intelligence techniques such as dimensionality reduction and neural and deep networks. You will gain hands-on experience through structured MATLAB assignments based upon adaptive acoustic noise cancellation, high-resolution frequency estimation from students' own physiological recordings (ECG), and frequency tracking in communications and smart grid applications.

Learning Outcomes

Upon successful completion of this module, you will be able to: 1. design and implement adaptive learning systems to suit particular requirements Adaptive learning systems 2. formulate and execute practical estimators in frequency domain Estimators 3. design non-linear and neural systems for machine intelligence Machine intelligence 4. test design ability on several practical case studies Case studies


Aspects of estimation theory: bias, variance, maximum likelihood and efficiency; Resolution and stability; time-bandwidth product; Classical spectral estimation: periodogram, averaging and Blackman-Tukey method; Parametric models: linear, rational transfer function, and non-linear models; Time-frequency and time-scale methods, order selection; Block and sequential parameter estimation techniques; Stochastic gradient type algorithms: least mean square and recursive least squares and Kalman filtering adaptive algorithm families; Multidimensional adaptive filters, widely linear estimators, dealing with data noncircularity; Blind equalization and source separation; Nonlinear online learning algorithms and their connection to neural networks and deep learning; Case studies.
Exam Duration: N/A
Exam contribution: 0%
Coursework contribution: 100%

Term: Spring

Closed or Open Book (end of year exam): N/A

Coursework Requirement:

Oral Exam Required (as final assessment): N/A

Prerequisite module(s): None required

Course Homepage: unavailable

Book List: