By Mandy Oakham
A Deakin professor has cracked the secret code of success with her research aimed at teaching machines how to think and learn.
Alfred Deakin Professor Svetha Venkatesh, who is the director of the Geelong-based Centre for Pattern Recognition and Data Analytics, has been awarded an Australian Laureate Fellowship.
To understand her research the professor explained that teaching machines how to think is like developing recipes for humans. Through analysing huge amounts of data the professor’s research is able to identify patterns in that data and then produce algorithms which can then be used to help them think in the same way as human beings would use a recipe to help them bake a cake.
For example using this model a machine such as a street camera could be given the “recipe“ to identify different behaviours as suspicious or normal. The security implications of this research are very obvious.
“Everybody is always interested in security issues,“ Professor Venkatesh said.
“It is very exciting to think that this kind of research is being done right here in Geelong.
“The possibilities of this research are truly limitless, and I hope the next stage of my work may use these principles to develop personalised medical analysis capabilities.“
The principles developed by her research have been applied into real world uses such as predicting suicide risk factors in patients presenting in emergency departments and an app which can help kids with autism gain basic learning skills.
Deakin Vice-Chancellor Professor Jane den Hollander congratulated the Deakin researcher on her award.
“We are particularly delighted for and proud of Professor Venkatesh, who has long played a leadership role in demonstrating to other Australian women just what is possible to achieve in the field of computing and pattern recognition globally.”
As a Professor of Computer Science, Professor Venkatesh is one of Australia’s leading experts in pattern analysis and her work has led to two successful start-ups.
