IEEE Computational Intelligence Magazine - August 2020 - 5

to do a second Master degree on nonsymbolic AI at the University of Edinburgh. That was my first encounter with
Connectionism and Genetic Algorithms,
but I also augmented my courses with
others in knowledge representation, natural language processing and automated
reasoning due to my passion for symbolism. My Master thesis combined symbolic and non-symbolic AI to solve a
problem in heat exchanger networks, to
which I was exposed after taking a
course from the department of chemical engineering. My thesis was awarded
the best AI thesis of the year at the
University of Edinburgh. During that
Master, I met a few people who truly
influenced my life, including my wife,
who kept growing my passion for linguistics and the philosophy of language.
By the end of that Master, my brain
was buzzing with new knowledge and
information that kept influencing my
way of thinking. I became a believer in
the importance of using the right technique for the right problem and avoiding biases when applying the same
technique to every problem. I started
building passion for philosophy and
language, although I did not enjoy any
of these topics before.
The third critical event in my life
which shaped my thinking and made
me who I am today was an offer I
received before I needed to return to
resume my lecturer position and continue my first uncompleted PhD at Cairo
University. I was awarded a PhD scholarship from Queensland University of
Technology (QUT), Brisbane, Australia
to work on data mining for the dairy
industry. The mere idea of quitting a
PhD right at the writing stage to start a
new one, and moving all the way to
Australia would sound inconceivable to
many. The risk and sacrifice were significant. Had I not taken that risk, my life
today would have been very different
and most likely, I would not have been
where I am currently.
With continuous fire in my belly and
a keen interest in improving myself, I
took the plunge and started a new PhD
from scratch on a topic that required
from me to study an area I knew noth-

Hussein Abbass with his family: Eleni, Adam and Zach.

ing about; that is, animal genetics which
was essential for the problem domain
where I needed to apply AI. During that
PhD, I learnt about rule extraction from
neural networks and invented dynamic
decision trees, where nodes have memories encoding Elman-type recurrent
equations. However, my scholarship was
supported from industry and I needed
to also design data mining and multistage multi-objective algorithms for the
optimization of genetic materials of
dairy cattle, which I did.
At the end of that PhD, I was ready
with a portfolio of mathematical and
algorithmic skills, and experience in
solving a wide range of problems across
so many diverse domains, from finance,
variety of engineering problems to animal genetics, and more! This was the
starting point for the twenty years that
followed and that brought me to where
I am now to become the Founding Editor-in-Chief of the IEEE Transactions
on Artificial Intelligence.
2. Can you tell us a little bit
abou1t your research work that
brought you to this appointment?

Over the last twenty years, and after
completing my PhD at QUT, I continued to work with multi-objective problems, both for optimization and
machine learning. I quickly realized the
value of, and invented Pareto-based
neural ensemble learning. I then had

two major application themes that funded the first 12 years post my PhD. One
was a set of technical projects in air traffic control and robotics with well-structured problems. The other set was sitting
at the opposite end, where problems are
wicked and ill-structured. I worked on
developing computational environments
and new algorithms for both. Optimization, simulation, machine learning and
knowledge based systems were the natural choice for the first theme. The second theme required me to find ways to
bring AI to wicked problems that mostly require understanding of complex
systems, system thinking, and game theory. This is when I proposed the Computational Red Teaming (CRT)
environment that I became known for.
CRT brought AI to wicked problems
to support long-term planning decisions, concept evaluations, strategy
development, which then became a
common technology in cyber security.
In the last 8 years of my career, I
started to bring the human to AI and
bring AI to the human. The skills I
developed after my first degree led me
naturally to appreciate cognitive engineering and design technical systems
that connect the human brain and psychophysiological data with AI. I developed
new architectures, such as cognitivecyber symbiosis, for seamless dynamic
coupling of humans and distributed
autonomous systems.

AUGUST 2020 | IEEE COMPUTATIONAL INTELLIGENCE MAGAZINE

IEEE Computational Intelligence Magazine - August 2020

Table of Contents for the Digital Edition of IEEE Computational Intelligence Magazine - August 2020