Junbin Gao

Biography

Junbin Gao is Professor of Big Data Analytics at the University of Sydney Business School. Prior to joining the University of Sydney in 2016, he was Professor in Computing from 2010 to 2016 and Associate Professor from 2005 to 2010 at Charles Sturt University (CSU). He was Senior Lecturer from Jan 2005 to July 2005 and Lecturer from Nov 2001 to Jan 2005 in the School of Mathematics, Statistics and Computer Science (now the School of Science and Technology) at University of New England (UNE). Between 1999 and 2001, he worked as a Research Fellow in the Department of Electronics and Computer Science at University of Southampton, England. Junbin Gao graduated from Huazhong University of Science and Technology (HUST) in 1982 with a Bachelor Degree in Computational Mathematics. He obtained his PhD from Dalian University of Technology in 1991. Between 1991 and 1993 he worked as a postdoctoral research fellow investigating wavelet applications at Wuhan University. He was appointed as an Associate Professor in July 1993 and promoted to Professor in October 1997 in Department of Mathematics of HUST. He was Guest Professor (2003-2006) in the State Key Lab of Information Engineering in Surveying, Mapping and Remote Sensing at Wuhan University, China; Guest Professor (2007-2010) in the School of Computer Science and Technology at Huazhong University of Science and Technology, China; Guest Professor (2008-2011) in the School of Computers at Guangdong University of Technology, China; and Visiting Professor (2012-2015) in Beijing Municipal Key Lab of Multimedia and Intelligent Software Technology at Beijing University of Technology. Until recently his major research interest has been machine learning and its application in data science, image analysis, pattern recognition, Bayesian learning & inference, and numerical optimization etc. He is the author of 260 academic research papers and two books. His recent research has involved new machine learning algorithms for big data in business. Prof Gao won two research grants in Discovery Project theme from the prestigious Australian Research Council (ARC).

Research Interest

Professor Junbin Gao began his research career by studying approximation theory and application of multivariate spline functions in numerical solutions for partial differential equations, continuing research work in wavelet applications in chemometrics, and becoming an outstanding researcher in machine learning, pattern recognition, Bayesian learning/inference, numerical optimisation and big data analytics in business. One of interesting examples in Junbin’s recent research is to propose matrix neural networks and apply it in longitudinal relational data in politics research, where he further develops it to tensorial recurrent neural network. In a series of papers from 2014 to 2017 Junbin Gao showed how to conduct data subspace clustering and dimensionality reduction on manifolds particularly for the abstract Grassmann manifolds. Much of this work has been joint with a number of international collaborators. Junbin Gao’s work prior to 2014 is on dimensionality reduction, which was funded by the Australian Research Council (ARC), and the success can be seen in a series of paper between 2005 and 2013 and the research was quoted by The Australian newspaper in 2012. More recently Junbin has focussed on designing machine learning algorithms for structural data such as tensor-valued data and manifold-valued data widely seen modern business and computer vision. In classical data analysis and machine learning algorithms, input data including manifold-valued data are generally regarded as or converted to vectorial data in a Euclidean space by ignoring useful prior information. However, for manifold-valued data, it is unclear how to extend those very powerful machine learning algorithms for vectorial data, such as the state-of-the-art Low Rank Representation models, onto general Riemannian manifolds due to loss of linearity structures over “curved” Riemannian manifolds. In recent years there has been great progress in the research of commonly used matrix manifolds such as the tensor manifold/covariance descriptor, Stiefel manifold, Multinomial Manifold, Grassmann manifold, Kendall Shape manifolds and Low Rank matrix manifold.  The core idea is to explicitly incorporate geometry of manifolds for the purpose of learning algorithm design, which brings advantages of improving accuracy and efficiency and reducing computational cost of conventional machine learning algorithms. So answering questions about manifold-valued data modelling forces us to consider how sufficiently using Riemannian properties of the well-known matrix manifolds to assist designing new learning algorithms for manifold-valued data analysis. There are applications in many areas, but Junbin Gao has a particular interest in what happens in international relation research and panel data in financial world, and also learning tasks in pattern analysis for computer vision tasks. © 2002-2017 The University of Sydney. Last updated: 8 September, 2017 ABN: 15 211 513 464. CRICOS number: 00026A. Phone: +61 2 9351 2222. Contact the University | Disclaimer | Privacy | Accessibility