I am genuinely interested in any topic that has to do with knowledge discovery from all kinds of data and this naturally extends to the intelligent systems design and implementation. Specifically, my research interests are the following:
- Machine learning.
- Software engineering.
Currently, wearable and mobile devices, Internet, and many types of research instruments (such as NGS) generate large amounts of data which require self-learning intelligent systems to extract previously unknown patterns that can be seen as the data summaries for better understanding of data-generative phenomena and processes through machine learning and predictive analytics.
For my PhD research project, I hypothesized, that it is possible to discover recurrent behaviors - that are the building blocks of software processes - from large amounts of publicly available software process artifacts, for example from software change records available at GitHub. Investigating this hypothesis, I have explored the possibility of data mining and machine learning techniques application for the analysis of software process and product artifacts measurements. So far, this exploratory study resulted in two concrete results:
- a novel interpretable time series classification technique called SAX-VSM that aids in the knowledge discovery from time series by finding and ranking of class-characteristic features.
- a confirmation of the research hypothesis through a number of case studies, where I show that significant features discovered by SAX-VSM do correspond to recurrent behaviors.
- an extension of SAX-VSM that is based on grammatical inference.
In my opinion, the successful software development is a result of a development discipline, that is a combination of appropriate tools, techniques, attitude, and habits. An effective combination of these enables teams and individuals not only to reach their goals efficiently, but to deliver reliable and performant software systems and components. The understanding of these "magic mixtures" is my next research interest.
Accidentally, years ago, I become involved in bioinformatics after I've got a research assistant job at ASGPB. My special thanks to Dr. Alam! Ten years later, I still work in the field. Probably the best thing about bioinformatics is the continuous collaboration with researchers from a variety of scientific disciplines - we have biologists, mathematicians, statisticians, computer scientists working within the same project. Moreover, it involves data mining, machine learning, and software engineering.