24 October 2022,
by Adil Ibrahim, Heriot-Watt University
Hi there! I’m Adil, a PhD student studying at Heriot-Watt University – Edinburgh. My research focuses on binary classification using artificial immune systems, and I use one of the primary immune algorithms, the Negative Selection Algorithm. It has always intrigued me to use artificial immune systems in data classification. My curiosity has always been piqued by how artificial immune systems can differentiate between self and nonself, depicting one of the main features of the biological immune system of our own.
For that to be achieved, there must be an affinity function that we could use so it works like the natural immune system. This affinity function has always been a challenge. In most cases, the suggested affinity functions did not work well, to the extent that many researchers concluded that the Negative Selection is not a reasonable classification algorithm.
I searched for a suitable affinity function and looked at the techniques used in bioinformatics. The protein sequence alignment schemes are used initially to measure the similarities between DNA, RNA, or protein to identify regions of similarity that may be a consequence of functional, structural, or evolutionary relationships between the sequences. So, I wondered if this could be the best affinity function to be used. So, I focused on the Negative Selection algorithm with bioinformatics protein sequence alignment schemes and used their mathematical techniques to build an affinity function required for the self and nonself discrimination in binary classification. The methods have been tested using datasets in the health domain to diagnose breast cancer and other datasets in different domains for comparisons.
My supervisor, Professor Nick Taylor, has told me about SICSA and the possibilities of getting support through their programs to attend a conference. So, as part of my research, I attended the 14th International Conference on Agents and Artificial Intelligence (ICAART 2022). Due to COVID-19, the conference was streaming online from Feb 3, 2022, to Feb 5.
The ICAART was an excellent opportunity for me as it brought together researchers, engineers and practitioners interested in the theory and applications in the areas of Agents and Artificial Intelligence.
The ICAART2022 was my first conference to attend as a speaker, and I needed to learn how it goes in such a contribution. The ICAART22 introduced me to many other researchers in different fields. The conference events and contributing to the discussions at the conference have expanded my experience considerably, bringing up several new research questions to explore since after.
It was a dream come true because the support I got from SICSA was highly fruitful, and I’m very grateful for that.