Log in

Login to your account

Username *
Password *
Remember Me
From Brain Science to Intelligent Machines

Haider Raza

Currently, I am working as a Post-Doctoral Research Officer for Machine Learning at The Farr Institute of Health Informatics Research, Swansea University Medical School, UK (July 2016 - Present). Previously, I have worked as a Post- Doctoral Research Assistant in Machine Learning for EEG-based Brain-Computer Interfacing at the Intelligent Systems Research Centre (ISRC), School of Computing and Intelligent Systems, Ulster University, UK (December 2015-June 2016).

My doctoral (PhD) dissertation was under the supervision of Professor Girijesh Prasad and Dr Hubert Cecotti and looked at adaptive learning for modelling for non-stationary systems and its application to EEG-based brain-computer interfaces. 

Research interest:

• Data Science and Data Analytics
• Machine Learning for Biomedical Signals and Health Care Datasets (Electronic Health Records)
• Dataset Shift/Change-Detection Method.
• Signal Processing: EEG, MEG, EMG decoding (feature extraction and features selection); blind source separation; dimensional reduction.
• Neural Engineering: Brain-Computer Interfaces for Communication and Stroke-Rehabilitation; Neuro-feedback.

PhD Project

PhD., Computer Science
Title: Adaptive Learning for Modelling Non-Stationary Systems in EEG-based Brain-Computer Interfacing
Supervisors: Professor Girijesh Prasad and Dr. Hubert Cecotti
Neural Systems and Neuro-Technology Team, Intelligent Systems Research Centre,
School of Computing and Intelligent Systems, Ulster University, Magee Campus,
Londonderry, BT48 7JL, Northern Ireland, United Kingdom (UK).

Key finding and contributions in PhD:

Contributed a covariate shift detection test for time-series data (Statistical test for covariate shift detection in Gaussian and non-Gaussian Datasets)
Designed an adaptive learning algorithm for handling non-stationarity in motor imagery-related EEG data (Supervised learning algorithm, covariate shift adaptation, signal processing, kernel methods, neural engineering, and time-series analysis)
Designed a transductive learning algorithm for covariate shift adaptation in EEG data (Bayesian inference, kernel method, Vapnik’s theory, neural engineering, and time-series analysis)

Contact: haider.raza at swansea dot ac dot uk