Research Fellow in Adversarial Machine Learning
School of Computing and Information Systems
Melbourne School of Engineering
Salary: $69,148* - $93,830 p.a. (*PhD entry Level A.6 $87,415) plus 9.5% superannuation
Postdoctoral research fellow open in the "Adversarial Machine Learning for Cybersecurity" project in the School of Computing and Information Systems at the University of Melbourne, Australia. The 1 year full-time position is funded by the Defence Science and Technology Group, with CSIRO/Data61.
This project aims to deliver new algorithmic and theoretical results on the robustness of machine learning systems, particularly reinforcement learning, in adversarial environments. Today Machine Learning and Statistics are employed in many technologies where participants have an incentive to game the system, for example network management, autonomous vehicles, medical devices, credit risk in finance, and smart utility grids. However little is known about how well state-of-the-art inference techniques fare when data is manipulated by a malicious adversary. Less is known about how to make learners robust to manipulation. This is particularly true for reinforcement learning. Key outputs of this project include publications reporting new algorithmic and theoretical ideas in adversarial learning, and demonstrations of adversarial reinforcement learning in software-defined networking.
The ideal candidate would be an enthusiastic researcher with strong skills in: machine learning algorithms (particularly but not limited to reinforcement learning and deep learning methods), optimisation & linear algebra, and programming. Experience with machine learning applied in security (e.g., intrusion detection) or game theory is a plus.
You will be part of a cross-institutional team spanning Melbourne and Swinburne Universities, DSTG and CSIRO/Data61, who collaboratively work on all aspects of the project, from concepts, theoretical underpinning, and design, through implementation and experimentation, to publication of results, curation of data, and maintenance of generated software. You will work side by side with other researchers in machine learning, game theory and computer security.
Being located in the School of Computing & Information Sciences within the Melbourne School of Engineering, you will be expected to be an active member of the School, collaborating with other researchers. You may undertake small amounts of teaching and research supervision directly related to your area of research, as required.
The Melbourne School of Engineering is strongly committed to supporting diversity and flexibility in the workplace. Applications for part-time or other flexible working arrangements will be welcomed and will be fully considered subject to meeting the inherent requirements of the position.
The University plan seeks to increase the diversity of the workforce and the representation of women in areas they have been traditionally under-represented. Consistent with this the School is seeking to increase the representation of women in the academic workforce across engineering disciplines. Under a Special Measure, under Section 12 (1) of the Equal Opportunity Act 2010 (Vic) the School is seeking to lift the representation of women from 20% in 2014 to at least 25% over the next 5 years, and strongly encourages applications from suitably qualified female candidates.
Close date: 31 Jan 2018
Position Description and Selection Criteria
For information to assist you with compiling short statements to answer the selection criteria, please go to http://about.unimelb.edu.au/careers/search/info/selection-criteria