Page 58 - University of Pretoria Research Review 2017

56
Artificial Intelligence – myth or reality?
Nelishia Pillay, Department of Computer Science
There are many expectations – some well-founded and others perhaps not – of the contributions that artificial intelligence can make to society and the quality of life, and in shaping a future that we have come to know as the Fourth Industrial Revolution.
For some time now artificial intelligence has played an important role in industry with applications ranging from the scheduling of jobs, vehicle routing, airplane landing, financial forecasting and, more recently, in media and broadcasting, smart cities, mining, and
in areas such as autonomous vehicles and energy consumption.
In education, artificial intelligence makes possible
the provision of individualised tutoring by means of intelligent tutoring systems and automated teaching assistants, assessment and data analytics in identifying individual learners’ learning difficulties. Artificial intelligence has also made inroads into society. In 1997, the artificial intelligence chess-playing programme Deep Blue beat world champion Gary Kasparov. Twenty years later, Alpha Go, a machine-learning Go player, beat number one Go player in the world, Ke Jie, at the Future of Go Summit. In the domain of the creative arts, artificial intelligence has produced a new generation
of artists and composers who use artificial intelligence techniques such as evolutionary algorithms to produce artwork and music compositions.
In South Africa artificial intelligence has contributed to growth in agriculture and the financial sector, as well as in combating crime by predicting incidents such as cable theft and rhino horn poaching. It is anticipated that artificial intelligence will contribute significantly to economic progress by 2030.
The Nature Inspired Computing Research Group (NICOG) at UP, led by Professor Nelishia Pillay, Head of the Department of Computer Science, focuses on using artificial intelligence in search techniques to solve a range of problems in scheduling, automatic programming, network intrusion detection, data mining, routing problems, packing problems, game playing and automated design, among others. These
search techniques take an analogy from nature, such as Darwin’s theory of evolution, in providing solutions to optimisation problems.
A current focus area of the group is hyper-heuristics, a fairly new optimisation technique aimed at providing more generalised solutions to real-world problems. The first book on hyper-heuristics, Hyper-Heuristics: Theory and Applications, co-authored by Professor Pillay, will be published by Springer in 2018. Professor Pillay’s current collaborations include the University of Nottingham in the UK, on hyper-heuristics, with MultiChoice on solving problems in broadcasting, and with the Department of Physics at UP and the Square Kilometre Array (South African Radio Astronomy Observatory) on scheduling problems. In 2017, Professor Pillay’s work on hyper- heuristics was published by the Institute of Electrical and Electronic Engineers (IEEE), Springer Science and Business Media, and in conference proceedings on Progress in Artificial Intelligence, also published by Springer.