|
Quantum Computing and Quantum Machine Learning
My current research deals primarily with Quantum Computing and Quantum Machine Learning.
In particular, I am working on new quantum machine learning algorithms for prediction and forecasting; applications of quantum technology;
as well as the creation of novel software environments and data visualisation techniques to support quantum model development.
Much of my effort focuses on Quantum Time Series Analysis.
Modeling of time series has received very little attention from the quantum community to date, hence,
much of the work in this area is without any prior research to refer to.
With my students and colleagues we explore the fundamental issues of time series quantum encoding,
suitability of quantum models for their processing, and extraction of useful insights from such models.
One of the most promissing and challenging area of time series analysis is Quantum Information Field Theory (QIFT),
which I actively collaborate on with Prof. Sebastian Zając.
QIFT based Quantum Data Science methods assist in discovering complex patterns and anomalies in data, which cannot be detected using established data analysis techniques.
Target applications of my research in quantum time series analysis include medical diagnosis (such as ECG),
machine condition monitoring (such as vibration analysis) and radar signal analysis (such as in electronic warfare).
However, business applications of this work are also being considered (such as stock market prediction and portfolio optimisation).
As the majority of investigated models involve variational quantum algorithms,
an important aspect of my research are the issues of quantum model trainability.
In particular, I investigate approaches to mitigating the emergence of barren plateaus -
the large flat areas in the cost function landscape, which negatively impact optimisation of large quantum circuits.
I advocate the need for the transition of quantum theory into practice,
hence I am always looking for new areas of this technology application, identification of its potential benefits to various stakeholders,
education of information professionals, as well as, elucidations of "obscure" quantum concepts to the general public.
|
Figure. Quantum higher-order regression fit of a time series (Click image)
Figure. Complex quantum fit of a time series with prediction (Click image)
Figure. Quantum Reservoir Computing analysis of chaotic data (Click image)
|
|
Classical Machine Learning and Data Visualisation
My past research activities were also devoted to AI applications in business, including
Machine Learning (with the focus on deep learning, predictive analytics and text analytics) and
Data Visualisation (featuring interactive data visualisation and immersive 3D data environments).
My earlier work also included data mining, expert systems, neural networks and natural language understanding.
My work in data visualisation also lead to projects, which explored human-computer interaction and relationships between human agents and technology,
including their decision making, communication, negotiation and collaboration, whether co-located or distributed.
|
Figure. Visualisation of a 3D data terrain that can be interacted with by gestures (Click image)
|
