Researchers from the Machine Learning Center at Georgia Tech (ML@GT) will present eight papers at the 24th International Conference on Artificial Intelligence and Statistics (AISTATS.) One paper by four Georgia Tech researchers, including ML@GT Associate Director Yao Xie, was accepted as an oral presentation, an accomplishment that only 3.1% of more than 1,500 accepted papers achieved.
“Me and my co-authors are honored to have our paper accepted as an oral presentation this year. ML@GT’s eight acceptances showcase the leading strength and breadth of machine learning research at Georgia Tech by our faculty and students,” said Xie, who is also an associate professor and Harold R. and Mary Anne Nash Early Career Professor in the H. Milton Stewart School of Industrial and Systems Engineering.
Accepted papers from Georgia Tech include:
- On the proliferation of support vectors in high dimensions by Daniel Hsu, Vidya Muthukumar, and Ji Xu
- Online Model Selection for Reinforcement Learning with Function Approximation by Jonathan N. Lee, Aldo Pacchiano, Vidya Muthukumar, Weihao Kong, and Emma Brunskil
- Feedback Coding for Active Learning by Gregory Canal, Matthieu Bloch, and Christopher Rozell
- Variational Autoencoder with Learned Latent Structure by Marissa Connor, Gregory Canal, and Christopher Rozell
- Learning to Defend by Learning to Attack by Haoming Jiang, Zhehui Chen, Yuyang Shi, Bo Dai, and Tuo Zhao
- Noisy Gradient Descent Converges to Flat Minima for Nonconvex Matrix Factorization by Tianyi Liu, Yan Li, Song Wei, Enlu Zhou, and Tuo Zhao
- Deep Fourier Kernel for Self-Attentive Point Processes by Shixiang Zhu, Minghe Zhang, Ruyi Ding, Yao Xie
- Goodness-of-Fit Test for Mismatched Self-Exciting Processes by Song Wei, Shixiang Zhu, Minghe Zhang, Yao Xie
The conference is in interdisciplinary gathering for researchers at the intersection of machine learning, statistics, artificial intelligence, and related areas. It will be held virtually April 13-15.