PhD Student

 

Statistical Machine Learning, NICTA Research School of Information Sciences & Engineering Australian National University Phone: +61 2 6125 1755 Email: jin {dot} yu {at} nicta.com.au A Brief Curriculum Vitae

 

About Me

I obtained my Bachelor's degree in Electrical Engineering from the Civil Aviation University of China and Master's degree in Artificial Intelligence from the Katholieke Universiteit Leuven in Belgium. I'm now a Ph.D. student in the RSISE at the Australian National University. I am sponsored by NICTA, where I am working with the Statistical Machine Learning group under the supervision of Dr. Nicol N. Schraudolph.

 

Research Interests

My main research interest is in stochastic (online) gradient methods for large-scale optimisation and machine learning. In the ANGie project, I am developing new rapid stochastic gradient methods for the ongoing adaptation of large, complex models to high-volume data streams.

 

Publications

2007

Nicol N. Schraudolph, Jin Yu, and Simon Günter. A Stochastic Quasi-Newton Method for Online Convex Optimization. In Proc. 11th Intl. Conf. Artificial Intelligence and Statistics (AIstats), pp. 433–440, Society for Artificial Intelligence and Statistics, San Juan, Puerto Rico, March 2007.
Details     [bib]    [pdf]   [mov] 

Silvia Richter, Douglas Aberdeen, and Jin Yu. Natural Actor-Critic for Road Traffic Optimisation. In Advances in Neural Information Processing Systems, The MIT Press, Cambridge, MA, 2007. Pre-proceedings version
Details     [bib]    [pdf] 

2006

Nicol N. Schraudolph, Jin Yu, and Douglas Aberdeen. Fast Online Policy Gradient Learning with SMD Gain Vector Adaptation. In Advances in Neural Information Processing Systems, pp. 1185–1192, The MIT Press, Cambridge, MA, 2006.
Details     [bib]    [pdf] 

 

Talks

Jin Yu, Nicol N. Schraudolph, and S.V.N. Vishwanathan. Online Limited-Memory Quasi-Newton Training of Support Vector Machines. Presented at Snowbird, San Juan, Puerto Rico, March 2007.
[pdf]    [mov] 

 

Previous Projects