Module Catalogue 2024/25

Pre Requisite Comment

Fundamental knowledge of signals and systems.
Basic MATLAB knowledge for the lab sessions.

Co Requisite Comment

N/A

Aims

To present emerging methods for the manipulation and analysis of single, multi-dimensional and random signals. And to conduct case studies in biomedical and healthcare applications.

To provide knowledge and in-depth explanation of discrete-time signal processing algorithms and approaches to measure deterministic and random signals in frequency domain.

To provide knowledge and recognition of supervised learning and unsupervised learning. And to design and demonstrate appropriate digital filters and techniques according to the application.

Outline Of Syllabus

Machine Perception Pattern Recognition Systems Design Cycle and Learning/Adaptation.

Supervised and Unsupervised Learning Techniques.

Data Dimensionality Reduction and Signal Enhancement Techniques.

Deterministic Signals, Transformation of Deterministic Time Signals into Frequency Domain.

Importance of Digital Filter in DSP, Realization of Filters, and Design of Digital Filters.

Random Sequences, Statistical Properties related to Random Sequences. Filtering Algorithms to Filter Random Sequences. Power Spectral Estimation.

Case Studies

Intended Knowledge Outcomes

At the end of this course students will acquire comprehensive knowledge to:

-       appreciate the key aspects of pattern recognition, detection, and classification [M2].

-       be familiar with supervised learning, unsupervised learning, and blind source separation [M1].

-       be aware of techniques for cluster analysis, such as the k-means algorithm and Gaussian Mixture Models (GMMs), Maximum Likelihood Estimation (MLE) and Expectation Maximization (EM) Algorithm [M2].

-       appreciate discrete-time signal and frequency transform methods [M1].

-       design finite impulse response, infinite impulse response, and adaptive filters [M3].

-       investigate random signals using statistical properties, such mean, variance, correlation, and can also determine spectrum [M1].

-       review an optimal filtering tool based on DSP-based application [M2].

Intended Skill Outcomes

At the end of course, students will be able to:

-       differentiate deterministic and random signals in time and frequency domain and will also be able to evaluate and compare the computational cost of different transform methods [M1].

-       work independently on Matlab/Simulink signal and analysis tool and classify the information and noise from given discrete signals [M2].

-       design and demonstrate digital filter to separate required/desired signal form signals with noise [M1].

-       design and employ spectral estimation to random signals [M1].

-       perform research investigation related to digital signal processing [M1].

-       perform case studies in biomedical and healthcare application [M4]

Teaching Rationale And Relationship

The face-to-face lectures and online synchronous sessions provide the fundamental concepts of the course.

The computer-based lab experience provides an opportunity to transform all theoretical learning into practical applications.

Case study provides research and studying state-of-the-art reading.

Assessment Rationale And Relationship

The examination allows students to demonstrate their ability to solve engineering problems focused on intelligent signal processing based on the knowledge and methodology presented in the course material. The laboratory report assesses technical writing skills and provides the opportunity for the students to apply programming skills to validate the theory taught on the course.

Formative analytical questions are provided as a homework to solve after the delivery of the related syllabus. The solutions are checked during the class and feedback is provided to gauge understanding and help students prepare for the summative assessments.

General Notes

N/A