Shiyu Liang's Homepage

Publications

• Robust Generalization in Overparameterized Neural Networks Using Explicit Regularization

  • Shiyu Liang, Ruoyu Sun and R. Srikant

  • Working paper, 2020

​

• The Role of Regularization in Overparameterized Neural Networks.

  • Siddhartha Satpathi, Harsh Gupta, Shiyu Liang and R Srikant.

  • IEEE Conference on Decision and Control, 2020

​

• Revisiting Landscape Analysis in Deep Neural Networks: Eliminating Decreasing Paths to Infinity

  • Shiyu Liang, Ruoyu Sun and R. Srikant, 2019

  • SIAM Journal of Optimization (submitted, current status: minor revision)

​

• FINE: A Framework for Distributed Learning on Incomplete Observations for Heterogeneous Crowdsensing Networks

  • Luoyi Fu, Songjun Ma, Lingkun Kong, Shiyu Liang and Xinbing Wang​

  • IEEE/ACM Transactions on Networking (ToN)​​, 2018

​

• Adding One Neuron Can Eliminate All Bad Local Minima

  • Shiyu Liang, Ruoyu Sun, Jason D. Lee and R. Srikant

  • Thirty-second Conference on Neural Information Processing Systems (NeurIPS), 2018

​

• Understanding The Loss Surface of Neural Networks for Binary Classification.

  • Shiyu Liang, Ruoyu Sun, Yixuan Li and R. Srikant.

  • International Conference on Machine Learning (ICML), 2018

​

• Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks.

  • Shiyu Liang, Yixuan Li and R. Srikant.

  • International Conference on Learning Representations (ICLR), 2018

​​

• Why Deep Neural Network For Function Approximation? 

  • Shiyu Liang and R. Srikant.

  • International Conference on Learning Representations (ICLR), 2017

​

• Are We Still Friends: Kernel Multivariate Survival Analysis.

  • Shiyu Liang, Ruotian Luo, Ge Chen, Songjun Ma, Le Song, Xiaohua Tian and Xinbing Wang. 

  • IEEE Global Communications Conference (GLOBECOM), 2014​​