Ruitong Huang

Independent Scholar, Canada

[first name] at ualberta.ca / rtong[last name] at gmail.com

I am about to join a finance company in London, UK. My research focuses on studying and designing provably efficient machine learning algorithms. The topics that I have worked on include Online Learning, Reinforcement Learning, Adversarial Learning/Robustness, and Statistical Learning Theory. My resume is here.

Experiences

• Research Team Lead, Borealis AI Lab (RBC Institute of Research)
  Nov. 2017 - Aug. 2020
  Research topics: reinforcement learning, adversarial robustness, and online learning
  
• Research Intern, Borealis AI Lab (RBC Institute of Research)
  Jul. 2017 - Nov. 2017
• Research Intern, Amazon Alexa
  Summer 2015

Education

• Ph.D student in Computing Science, University of Alberta
  Sept. 2010 - Sept. 2017
  Supervisors: Dale Schuurmans & Csaba Szepesvári
• Master of Mathematics in Computer Science, University of Waterloo
  Sept. 2008 - Jun. 2010
  Supervisor: Mark Giesbrecht
• Bachelor of Mathematics, University of Science and Technology of China (USTC)
  Sept. 2004 - Jul. 2008
  Supervisor: Junming Xu

Publications

• Towards minimax robust estimation via GANs
  K. Wu, GW. Ding, R. Huang, Y. Yu
  International Conference on Artificial Intelligence and Statistics (AISTATS), 2020
• Max-margin adversarial (MMA) training: Direct input space margin maximization through adversarial training
  GW. Ding, Y. Sharma, K. Lui, R. Huang
  International Conference on Learning Representations (ICLR), 2020.
• Maximum Entropy Monte-Carlo Planning
  C. Xiao, J. Mei, R. Huang, D. Schuurmans, M. Müller
  Neural Information Processing Systems (NIPS), 2019.
• On principled entropy exploration in policy optimization
  C. Xiao, J. Mei, R. Huang, D. Schuurmans, M. Müller
  International Joint Conference on Artificial Intelligence (IJCAI), 2019.
• On the sensitivity of adversarial robustness to input data distributions
  GW. Ding, K. Lui, X. Jin, L. Wang, R. Huang
  International Conference on Learning Representations (ICLR), 2019.
• Dimensionality reduction has quantifiable imperfections: two geometric bounds
  K. Lui, GW. Ding, R. Huang, R. McCann
  Neural Information Processing Systems (NIPS), 2018.
• Improving GAN training via binarized representation entropy (BRE) regularization
  Y. Cao, GW. Ding, K. Lui, R. Huang
  International Conference on Learning Representations (ICLR), 2018.
• Structured Best Arm Identification with Fixed Confidence.
  R. Huang, M. Ajallooeian, C. Szepesvári, M. Müller
  International Conference on Algorithmic Learning Theory (ALT), 2017.
• Following the Leader and Fast Rates in Linear Prediction: Curved Constraint Sets and Other Regularities
  R. Huang, T. Lattimore, A. György, C. Szepesvári
  Neural Information Processing Systems (NIPS), 2016.
  Journal version accepted to the Journal of Machine Learning Research (JMLR), 2017.
• Anchored Speech Detection
  R. Maas, S.H.K. Parthasarathi, B. King, R. Huang, B. Hoffmeister
  Interspeech, 2016.
• Revise Saturated Activation Functions
  B. Xu, R. Huang, M. Li
  Workshop Track, International Conference on Learning Representations (ICLR), 2016.
• Learning with a Strong Adversary
  R. Huang, B. Xu, D. Schuurmans, C. Szepesvári
  CoRR, abs/1511.03034.
• Easy Data for Independent Component Analysis
  R. Huang, A. György, C. Szepesvári
  Workshop “Learning Faster from Easy Data II” in Neural Information Processing Systems(NIPS), 2015.
• Deterministic Independent Component Analysis.
  R. Huang, A. György, C. Szepesvári
  International Conference on Machine Learning (ICML), 2015.
• A Finite-Sample Generalization Bound for Semiparametric Regression: Partially Linear Models.
  R. Huang, C. Szepesvári
  The 17th International Conference on Artificial Intelligence and Statistics (AISTATS), 2014.
• Generalization Bounds for Partially Linear Models.
  R. Huang, C. Szepesvári
  Special session on Theory of Machine Learning, International Symposium on Artificial Intelligence and Mathematics (ISAIM), 2014.
• Convex sparse coding, subspace learning, and semi-supervised extensions.
  X. Zhang, Y. Yu, M. White, R. Huang, D. Schuurmans
  In Annual Conference on Artificial Intelligence (AAAI), 2011.

Theses

• Instance-dependent analysis of learning algorithms. (PhD Thesis) pdf
• Decomposition of Finite-Dimensional Matrix Algebras over $F_q(y)$. （Master Thesis）pdf
• Decreasing the Diameter of Bounded Degree Graphs. (Bacherlor Thesis)

Awards and Honors

PhD outstanding thesis award (runner up), 2017
Provost Doctoral Entrance Award, University of Alberta, 2010&2011
Graduate Entrance Scholarship, University of Waterloo, 2008
Outstanding Undergraduate Thesis Award, USTC, 2008
National Scholarship (honored by China Ministry of Education), USTC, 2007
China Aerospace Science and Technology Corporation Scholarship, USTC, 2006

Services

Reviewer:

• Conferences: AISTATS, ICML, ALT, COLT, NIPS, IJCAI, ICLR, AAAI, ACML, UAI
• Journals: Journal of Machine Learning Research, Annals of Applied Statistics, Machine Learning, Neurocomputing, Neural Networks

Teachings

Teaching assistant at the University of Alberta for the following courses:
CMPUT 101: Introduction to Computing
CMPUT 210: Codes, Codemakers, Codebreakers: An Introduction to Cryptography
CMPUT 272: Formal Systems and Logic in Computing Science
CMPUT 304: Algorithms II
CMPUT 466/551: Machine Learning