The field of machine learning-based decision support is about to take a leap forward
She’s a world-leading researcher in areas of artificial intelligence (AI) like fuzzy transfer learning, concept drift and decision support systems, among others. And over the next five years, UTS Distinguished Professor Jie Lu is set to make Australia a global leader in another AI discipline: autonomous machine learning for decision support.
Professor Lu, the Director of the UTS Centre for Artificial Intelligence and a fellow of both the Institute of Electrical and Electronic Engineers and the International Fuzzy Systems Association, is about to embark on a five-year Australian Laureate Fellowship. It’s one of the most prestigious research appointments on offer, open to leading researchers around the globe in an effort to build world-class AI research capacity in Australia.
The first-ever UTS academic to be awarded the Laureate Fellowship, Professor Lu will receive $3.3 million in funding from the ARC, plus $2.3 million additional funding from UTS. She’ll use the funding to develop a series of innovative frameworks, models, algorithms and applications that will enable machine learning systems to support decision making in very complex (uncertain, dynamic, massive) situations.
“Decision support systems are used by various organisations to assist with their decision making,” Lu says.
“However, current decision support system research is struggling to keep pace as organisations increasingly seek to use Big Data to support more effective, intelligent and accurate decision-making.
“As such, there is an urgent need to co-opt machine learning into decision-making processes in order to establish advanced learning-based decision support systems.”
Uncertainty refers to things like ambiguous or unlabelled data values or the relationships of many data streams. Dynamic data contains patterns that change over time, requiring decision-makers to react quickly to gain insights and advantages from the data, while massive refers to data volume – from a few to millions of data records, dozens of domains, and/or massive data streams.
“Currently, machine learning-based decision technology cannot deal with these three complex situations. Though massive data are now available for use to support business decision-making, is beyond the capacity of existing machine learning and decision-making tools,” Professor Lu says.
“What I’m going to deal with is to develop autonomous machine learning algorithms to support decision-making when you have data in dozens of domains, a hundred data streams and various uncertainties in data – that’s a big data environment.”
Professor Lu’s algorithms will have the potential to transform any industry that relies on data-driven decision making, from food to health care and everything in between.
For cybersecurity companies, for example, quickly identifying ‘drift’, or changes to regular data patterns in data streams, could indicate an imminent cyber-attack. In the transport industry, real-time data from self-driving cars could better predict the risk of accidents based on factors like driver behaviours and road conditions.
The impacts on both fundamental research and industry, then, are potentially limitless – which is why Professor Lu’s project could put Australia front and centre of the global AI research field.
“AI, particularly machine learning, is having a transformational impact on many areas of science and of the economy,” Lu says. “This project will directly develop Australia’s globally competitive advantage in AI technologies and its applications.”