Welcome to Dr. Amir Hussain's PhD Research Opportunities page
Link to my home page
Please email me ( ahu@cs.stir.ac.uk ), if you are
interested in pursuing any of the
research ideas/areas below, OR indeed with any of your own ideas !!
Project
1: Novel Computational Intelligence Methods for Immune System Modeling and
Analysis
Principal
Supervisor: Dr. A. Hussain (Computing Science,
Additional Supervisor: Dr. Rachel Norman (Maths,
Outline:
Computing Science and Maths
at
In this PhD project, the research student would
have the opportunity to work on one (or both) of the following exciting
research areas (though I am open to other new ideas from potential
applicants!):
(i) To develop
a novel model of autoimmunity using artificial immune system techniques.
The application of a MATLAB toolbox based on recently developed artificial
immune systems techniques could be employed by the student for this project.
The knowledge developed in this project will be useful for understanding and
manipulating immune responses to infectious organisms. If the immune response
is turned on inappropriately, then autoimmunity ensues. Thus,
the decisions that the immune system makes to be turned on or off in
appropriate circumstances have a major impact on health and wellbeing.
Therefore, understanding these processes is a key challenge in immunology. The
input of both mathematical and computational modelling
can potentially make a significant contribution to our understanding of
autoimmune diseases.
(ii) To build a novel model of the natural
immune system as a whole using stochastic dynamics of interacting populations.
The application of process algebra based tools (already developed and currently
being used for related research in the department) will be investigated for
this project. The eventual aim of the project is to help us understand how the
human immune system maintains its diversity of millions of lymphocyte populations,
how populations of naive and memory cells are maintained, to determine the
turnover rates of various lymphocyte populations, and to understand the
possible homeostatic mechanisms regulating lymphocyte population sizes.
Project
2: Novel Computational Intelligence Techniques for Real-world Problem
Solving
Principal
Supervisor: Dr. A. Hussain (Natural Computing Research Group,
Possible External Supervisors: Dr. Calum MacRae (
Outline:
Currently, there is considerable interest in the
development of novel computational intelligence techniques and their
applications to solving practical problems e.g. in the medical, defense or
business (such as financial and telecommunications) industries.
Example PhD problems in the above areas, include:
(i) Computational intelligence based
decision-support methods for assisting medical practitioners (e.g. to surgeons
on when to perform appendicitis operations, cardiovascular preventative care,
chemotherapy symptom management etc.).
(ii) Computational Intelligence based advanced modeling and
analysis techniques for supporting financial decision making (e.g. in credit/fraud
scoring applications to prevent fraud, financial forecasting etc.). For
further (background) information, see related (downloadable papers) from: http://www.cs.stir.ac.uk/~ahu/Publications.htm (or
email me for further details: ahu@cs.stir.ac.uk)
Main idea behind the above
projects, is that since the underlying decision-spaces (e.g.
associated with medical diagnosis, financial fraud etc.) are highly
non-linear, this warrants the application of non-linear computational
intelligence techniques such as, artificial neural networks, expert systems,
evolutionary computation, as well as hybrid (e.g. neuro-fuzzy,
wavenet) techniques etc.
Project
3: Auditory processing modeling for future improved binaural hearing-aids
& speech recognition/processing applications
Principal Supervisor: Dr. A Hussain (Hearing
Research Lab,
Outline:
Dr. Hussain's Hearing
Lab in Stirling is a lead member (Vice-Chair and
Grant Holder) of the large (€0.5million) European Science Foundation
(ESF) funded European Research Network (COST-2102) that is concerned with the
development of new computational and mathematical models and algorithms to
drive the implementation of the next generation of telecommunication services
such as remote health monitoring systems, interactive dialogue systems, and
intelligent avatars. For further details: see: http://cost2102.cs.stir.ac.uk
One of the priority (PhD) research areas for
the Lab is concerned with the design, development, implementation and
subjective assessment of new multi-sensor sub-band adaptive DSP algorithms
inspired by early auditory processing features e.g
within-band, cross-band, band-selective non-linear strategies (such as neural,
Higher-order Statistics (HOS), non-linear Independent Component Analysis (ICA),
non-linear masking etc.).
Recent preliminary quantitative & qualitative
results with proposed methods warrant further investigation, and development of
a hearing-aid prototype (using a suitable DSP/FPGA hardware implementation
platform), for which further (PhD-level)
research can be pursued. Proposed speech enhancement schemes are generic
in that they offer the possibility of 'binaural unmasking' out with the human
body, providing signals of improved Signal to Noise Ratio and intelligibility
to the better human ear, a speech recognizer, conventional aid or cochlear
implant processor.
Other PhD research opportunities in this
exciting area can include: development of novel multi-modal
(audio-visual) processing methods to improve the next generation of
telecommunications services (including intelligent avatars, remote health
monitoring and interactive dialogue systems), and development of new
speech processing (i.e. enhancement, analysis, synthesis and recognition)
methods for foreign languages (Arabic, Urdu, etc.)
For more (background) details see related (downloadable)
papers here: http://www.cs.stir.ac.uk/~ahu/Publications.htm (or
email me for further details: ahu@cs.stir.ac.uk
)
Project
4: Neurobiologically inspired Cognitive Modeling and
Control for Complex real-world Systems
Principal
Supervisor: Dr. A Hussain (Intelligent Control Systems Lab,
Possible External Supervisor: Prof. Kevin Gurney
(Computational Neuroscience,
Outline:
The aim is to exploit emerging key common functional principles between
intelligent control theory and the vertebrate brain in order
to develop new neurobiologically motivated cognitive
control algorithms (that deploy integrated sensing, learning modeling, and
action selection / decision making capabilities) for industrial & medical
applications (such as, robotic control, autonomous vehicle control including
automated highway systems, insulin regulation of blood sugar &
diabetes etc.).
For more details (on background of neurobiologically inspired control algorithms and
applications): see recent paper (Book Chapter) by Hussain
and Gurney et al. downloadable from here: http://www.cs.stir.ac.uk/~ahu/ICANN08_final-paper.pdf .
Other related (downloadable) papers can be found here: http://www.cs.stir.ac.uk/~ahu/Publications.htm (or
email me for further details: ahu@cs.stir.ac.uk
)
Project
No. 5: Common Sense Computing for Next Generation Intelligent Web Applications
Principal
Supervisor: Dr. A. Hussain (Natural Computing Research Group,
Possible External Supervisor: Dr. C. Havasi (MIT Media Lab,
Outline:
Commonsense reasoning is the branch of Artificial
intelligence concerned with replicating human thinking. Commonsense computing
can enable web services to be more intuitive and people-friendly. Applied in
conjunction with Natural Language Processing technologies, it dramatically
enhances HCI, which is a key element for many applications especially in the
fields of e-commerce, e-tourism and e-health.
This project will involve using common sense
knowledge - including intelligent agents, natural language processing,
statistical machine learning, semantic data mining and multi-modal HCI
methods - to enable development of
state-of-the-art web applications in collaboration with MIT Media Lab and SiteKit Labs. One of the research objectives will
be be to develop a novel auto categorization
prototype tool for documents for knowledge and content management
applications (by exploring enhancements to MIT's existing Divisi software prototype and the ConceptNet
knowledge base). The project is likely to include a
regular (paid) industrial placement with SiteKit
Labs (at the serene and picturesque Isle of Skye in
For more (background) details on a related EPSRC
funded industrial PhD project, see: http://labs.sitekit.net/intelligentweb (or
email me for further details: ahu@cs.stir.ac.uk
)
Project
No. 6: Non-linear Computational Intelligence based Signal Processing algorithms for challenging real world
applications
Principal Supervisor: Dr. A. Hussain (Natural
Computing Research Group,
Possible External Supervisor: Prof. T. Durrani (Head, Centre of Excellence in Signal Image
Processing,
Outline:
Prior and ongoing research by Dr Hussain has highlighted and demonstrated the potential of
advanced machine learning techniques (including the use of feedforward
Support Vector Machines) for solving a range of challenging real world
problems, for example, neural networks have been applied for target bearing
estimation, mobile location estimation in cellular networks, multi-sensor adaptive beamforming
in reverberant non-stationary environments, recurrent neural networks for
non-linear prediction and adaptive equalization for combating multipath and
co-channel interference in mobile cellular systems (including adaptive blind
equalization using computationally efficient Higher Order Statistics based
approaches). Significant relevant extensions have been made to machine learning
theory and promising experimental results have been obtained using both
simulated and real data.
The aim of this PhD project is to develop new
computational intelligence/machine learning based signal processing algorithms
for real world problem solving. Of particular interest is to develop
new signal processing algorithms for:
(i) High Resolution Localization: The aim will be to develop novel robust signal processing based
multiple-target detection and localization techniques that can yield faster
and more accurate target bearing or direction of arrival (DoA) estimates compared to
conventional techniques, particularly for the case of closely spaced targets.
(ii) Broadband signal separation: The new techniques will also be extended to deal with the
challenging case of detecting and localizing multiple simultaneous targets from
broadband signals received at a sensor array. These technologies are also
essential in communications systems.
In addition, the proposed signal processing
research will lead to the detection and localization of non-stationary
moving targets (with time varying DoAs), in the
presence of multipath effects.
The performance of the developed algorithms will be
compared to state-of-the-art direction finding approaches and quantified using
simulation case studies including measured radar data involving closely spaced emitters, highly correlated non-stationary
signals and diffuse multipath effects (representing a challenging low-angle
tracking radar environment).
For further (background) information, see related
(downloadable papers) from: http://www.cs.stir.ac.uk/~ahu/Publications.htm (or
email me for further details: ahu@cs.stir.ac.uk )