Welcome

Bo DAI

Assistant Professor, Georgia Institute of Technology

Staff Research Scientist, Google Brain

I have moved to new homepage (https://bo-dai.github.io/).

This website is no longer actively updated.

Email: bodai AT google.com (for Google related work)

bodai AT cc.gatech.edu (for general academic work)

Twitter: daibond_alpha

Google Scholar

Short Bio

I am an assistant professor at Georgia Tech and a staff research scientist in Google Brain. I obtained my Ph.D. from Computational Science and Engineering at Georgia Tech.

My principal research interest lies on principled machine learning for structured data. I am particularly focusing on developing effective statistical models and efficient algorithms which learns from a massive volume of complex, structured, uncertain and high-dimensional data, e.g., distributions, structures, dynamics, and so on.

My research work includes three major themes:

Reinforcement learning: design effective algorithms by exploiting the intrinsic structures in the uncertain dynamics for automatic decision making.
Data-driven algorithm design: improve the algorithms, e.g., sampling, searching and planning, by leveraging empirical experiences.
Large-scale nonparametric machine learning: develop efficient algorithms for machine learning methods, especially nonparametric methods, to handle hundreds of millions of data.

More information can be found in my Google AI page.

News

2023/07: Invited talk at ICSP 2023, Davis.
2023/05: One paper have been accepted to UAI2023.
2023/05: One paper have been accepted to ICML2023.
2023/05: Invited talk at SIAM OP23, Seattle.
2023/05: Invited talk at AFOSR Workshop - Topics at the Intersection of Deep Learning and Computational Nonlinear Control.
2023/04: Invited talk at CMU, Pittsburgh.
2023/02: Serve as a Senior AC for NeurIPS 2023.
2023/01: Four papers have been accepted to ICLR2023.
2023/01: Two papers have been accepted to AISTASTS2023.
2022/11: Guest lecture at CS285, Deep Reinforcement Learning, University of California, Berkeley.
2022/09: Three papers have been accepted to NeurIPS2022.
2022/09: Invited talk at Amii, University of Alberta.
2022/08: Invited talk at Nanjing University.
2022/07: Invited talk at Chinese Academy of Science, Institute of Automation.
2022/07: Invited talk at MIT-IBM Watson AI Lab.
2022/06: Invited talk at Georgia Tech.
2022/05: Three papers have been accepted to ICML2022.
2022/05: One paper has been accepted to UAI2022.
2022/05: One paper has been accepted to KDD2022.
2022/04: Invited talk at Harvard University.
2022/01: Two papers have been accepted to ICLR2022.
2022/01: Two papers have been accepted to AISTASTS2022.
2021/12: I am now an Action Editor for TMLR.
2021/11: Invited talk at Workshop on Urban Mobility Simulation and Optimization.
2021/10: Invited talk at Facebook AI Research.
2021/09: Four papers have been accepted to NeurIPS2021.
2021/08: One paper, applying DICE for dialogue evaluation, has been accepted to EMNLP2021 as oral presentation.
2021/07: I am co-organizing The First Workshop on Evaluations and Assessments of Neural Conversation Systems (EANCS), co-located with EMNLP2021.
2021/07: Invited talk at Workshop on Reinforcement Learning at ICML2021.
2021/06: Lecture at EPFL&ETHZ Summer School: Reconciling Reinforcement Learning: Optimization, Generalization, and Exploration
2021/06: Guest lecture at University of California, Los Angeles.
2021/05: Four papers have been accepted to ICML2021.
2021/01: Our DICE family for Offline RL [1, 2, 3, 4] is reported in Google AI Year Review.
2021/01: One paper has been accepted to AISTATS2021.
2020/12: Invited talk at Purdue University.
2020/11: Invited talk at INFORMS Annual Meeting, Reinforcement Learning Theory.
2020/11: Invited talk at University of California, Los Angeles.
2020/10: Invited talk at RIKEN.
2020/09: Seven papers have been accepted to NeurIPS2020.
2020/09: Invited talk at Deep Reinforcement Learning at Simons Institute.
2020/06: Three papers have been accepted to ICML2020.
2020/05: Invited talk at SIAM Conference on Optimization 2020.
2020/04: Invited talk at Facebook AI Research.
2020/01: Invited talk at DeepMind Edmonton.

Selected Publications

Reinforcement Learning

Jincheng Mei*, Yue Gao*, Bo Dai, Csaba Szepesvári, and Dale Schuurmans. "Leveraging Non-uniformity in First-order Non-convex Optimization". The 38th International Conference on Machine Learning (long talk, ICML'2021).
Bo Dai*, Ofir Nachum*, Yinlam Chow, Lihong Li, Csaba Szepesvari, Dale Schuurmans. "CoinDICE: Off-Policy Confidence Interval Estimation". Neural Information Processing Systems (Spotlight, NeurIPS'2020).
Jincheng Mei, Chenjun Xiao, Bo Dai, Lihong Li, Csaba Szepesvari, Dale Schuurmans. "Escaping the gravitational pull of softmax". Neural Information Processing Systems (Oral, NeurIPS'2020).
Ruiyi Zhang*, Bo Dai*, Lihong Li, Dale Schuurmans. "GenDICE: Generalized Offline Estimation of Stationary Values". The 8th International Conference on Learning Representations (Oral, ICLR'2020).
Ofir Nachum*, Yinlam Chow*, Bo Dai, Lihong Li. "DualDICE: Behavior-Agnostic Estimation of Discounted Stationary Distribution Corrections". Neural Information Processing Systems (Spotlight, NeurIPS'2019). [CODE]
Bo Dai, Albert Shaw, Lihong Li, Lin Xiao, Niao He, Zhen Liu, Jianshu Chen, Le Song. "SBEED: Convergent Reinforcement Learning with Nonlinear Function Approximation". The 35th International Conference on Machine Learning (long talk, ICML'2018).
Bo Dai, Niao He, Yunpeng Pan, Byron Boots, and Le Song. "Learning from Conditional Distributions via Dual Embeddings". The 20th International Conference on Artificial Intelligence and Statistics (AISTATS'2017).

Learning to Design Algorithm

Hanjun Dai, Rishabh Singh, Bo Dai, Charles Sutton, Dale Schuurmans. "Learning Discrete Energy-based Models via Auxiliary-variable Local Exploration". Neural Information Processing Systems (NeurIPS'2020).
Binghong Chen*, Bo Dai*, Qinjie Lin, Guo Ye, Han Liu, Le Song. "Learning to Plan in High Dimensions via Neural Exploration-Exploitation Trees". The 8th International Conference on Learning Representations (Spotlight, ICLR'2020).
Bo Dai*, Zhen Liu*, Hanjun Dai*, Niao He, Arthur Gretton, Le Song, Dale Schuurmans. "Exponential Family Estimation via Adversarial Dynamics Embedding". Neural Information Processing Systems (NeurIPS'2019). [CODE]
Hanjun Dai, Zornitsa Kozareva, Bo Dai, Alexander Smola and Le Song. "Learning Steady-States of Iterative Algorithms over Graphs". The 35th International Conference on Machine Learning (ICML'2018). [CODE]
Hanjun Dai, Bo Dai and Le Song. "Discriminative Embeddings of Latent Variable Models for Structured Data". The 33th International Conference on Machine Learning (ICML'2016). [CODE]

Scalable Nonparametric Methods

Bo Dai, Niao He, Hanjun Dai and Le Song. "Provable Bayesian Inference via Particle Mirror Descent", The 19th International Conference on Artificial Intelligence and Statistics (Best Student Paper Award, full oral presentation, AISTATS'2016).
Bo Dai, Bo Xie, Niao He, Yingyu Liang, Anant Raj, Maria-Florina Balcan and Le Song. "Scalable Kernel Methods via Doubly Stochastic Gradients", Neural Information Processing Systems (NIPS'2014). [CODE]