Timezone: »
To address the challenge of backpropagating the gradient through categorical variables, we propose the augment-REINFORCE-swap-merge (ARSM) gradient estimator that is unbiased and has low variance. ARSM first uses variable augmentation, REINFORCE, and Rao-Blackwellization to re-express the gradient as an expectation under the Dirichlet distribution, then uses variable swapping to construct differently expressed but equivalent expectations, and finally shares common random numbers between these expectations to achieve significant variance reduction. Experimental results show ARSM closely resembles the performance of the true gradient for optimization in univariate settings; outperforms existing estimators by a large margin when applied to categorical variational auto-encoders; and provides a "try-and-see self-critic" variance reduction method for discrete-action policy gradient, which removes the need of estimating baselines by generating a random number of pseudo actions and estimating their action-value functions.
Author Information
Mingzhang Yin (University of Texas at Austin)
Yuguang Yue (University of Texas at Austin)
Mingyuan Zhou (University of Texas at Austin)
Related Events (a corresponding poster, oral, or spotlight)
-
2019 Oral: ARSM: Augment-REINFORCE-Swap-Merge Estimator for Gradient Backpropagation Through Categorical Variables »
Thu. Jun 13th 11:35 -- 11:40 PM Room Grand Ballroom
More from the Same Authors
-
2022 : Optimization-based Causal Estimation from Heterogenous Environments »
Mingzhang Yin · Yixin Wang · David Blei -
2023 Poster: Prototype-oriented unsupervised anomaly detection for multivariate time series »
yuxin li · Wenchao Chen · Bo Chen · Dongsheng Wang · Long Tian · Mingyuan Zhou -
2023 Poster: Bayesian Progressive Deep Topic Model with Knowledge Informed Textual Data Coarsening Process »
Zhibin Duan · Xinyang Liu · Yudi Su · Yishi Xu · Bo Chen · Mingyuan Zhou -
2023 Poster: Learning to Jump: Thinning and Thickening Latent Counts for Generative Modeling »
Tianqi Chen · Mingyuan Zhou -
2023 Poster: POUF: Prompt-Oriented Unsupervised Fine-tuning for Large Pre-trained Models »
Korawat Tanwisuth · Shujian Zhang · Huangjie Zheng · Pengcheng He · Mingyuan Zhou -
2022 Poster: Deep Variational Graph Convolutional Recurrent Network for Multivariate Time Series Anomaly Detection »
Wenchao Chen · Long Tian · Bo Chen · Liang Dai · Zhibin Duan · Mingyuan Zhou -
2022 Poster: Bayesian Deep Embedding Topic Meta-Learner »
Zhibin Duan · Yishi Xu · Jianqiao Sun · Bo Chen · Wenchao Chen · CHAOJIE WANG · Mingyuan Zhou -
2022 Spotlight: Bayesian Deep Embedding Topic Meta-Learner »
Zhibin Duan · Yishi Xu · Jianqiao Sun · Bo Chen · Wenchao Chen · CHAOJIE WANG · Mingyuan Zhou -
2022 Spotlight: Deep Variational Graph Convolutional Recurrent Network for Multivariate Time Series Anomaly Detection »
Wenchao Chen · Long Tian · Bo Chen · Liang Dai · Zhibin Duan · Mingyuan Zhou -
2022 Poster: Regularizing a Model-based Policy Stationary Distribution to Stabilize Offline Reinforcement Learning »
Shentao Yang · Yihao Feng · Shujian Zhang · Mingyuan Zhou -
2022 Spotlight: Regularizing a Model-based Policy Stationary Distribution to Stabilize Offline Reinforcement Learning »
Shentao Yang · Yihao Feng · Shujian Zhang · Mingyuan Zhou -
2021 Poster: Bayesian Attention Belief Networks »
Shujian Zhang · Xinjie Fan · Bo Chen · Mingyuan Zhou -
2021 Spotlight: Bayesian Attention Belief Networks »
Shujian Zhang · Xinjie Fan · Bo Chen · Mingyuan Zhou -
2021 Poster: Sawtooth Factorial Topic Embeddings Guided Gamma Belief Network »
Zhibin Duan · Dongsheng Wang · Bo Chen · CHAOJIE WANG · Wenchao Chen · yewen li · Jie Ren · Mingyuan Zhou -
2021 Poster: ARMS: Antithetic-REINFORCE-Multi-Sample Gradient for Binary Variables »
Alek Dimitriev · Mingyuan Zhou -
2021 Spotlight: ARMS: Antithetic-REINFORCE-Multi-Sample Gradient for Binary Variables »
Alek Dimitriev · Mingyuan Zhou -
2021 Spotlight: Sawtooth Factorial Topic Embeddings Guided Gamma Belief Network »
Zhibin Duan · Dongsheng Wang · Bo Chen · CHAOJIE WANG · Wenchao Chen · yewen li · Jie Ren · Mingyuan Zhou -
2020 Poster: On hyperparameter tuning in general clustering problemsm »
Xinjie Fan · Yuguang Yue · Purnamrita Sarkar · Y. X. Rachel Wang -
2020 Poster: Thompson Sampling via Local Uncertainty »
Zhendong Wang · Mingyuan Zhou -
2020 Poster: Bayesian Graph Neural Networks with Adaptive Connection Sampling »
Arman Hasanzadeh · Ehsan Hajiramezanali · Shahin Boluki · Mingyuan Zhou · Nick Duffield · Krishna Narayanan · Xiaoning Qian -
2020 Poster: Recurrent Hierarchical Topic-Guided RNN for Language Generation »
Dandan Guo · Bo Chen · Ruiying Lu · Mingyuan Zhou -
2019 Poster: Convolutional Poisson Gamma Belief Network »
CHAOJIE WANG · Bo Chen · SUCHENG XIAO · Mingyuan Zhou -
2019 Poster: Locally Private Bayesian Inference for Count Models »
Aaron Schein · Steven Wu · Alexandra Schofield · Mingyuan Zhou · Hanna Wallach -
2019 Oral: Convolutional Poisson Gamma Belief Network »
CHAOJIE WANG · Bo Chen · SUCHENG XIAO · Mingyuan Zhou -
2019 Oral: Locally Private Bayesian Inference for Count Models »
Aaron Schein · Steven Wu · Alexandra Schofield · Mingyuan Zhou · Hanna Wallach -
2018 Poster: Inter and Intra Topic Structure Learning with Word Embeddings »
He Zhao · Lan Du · Wray Buntine · Mingyuan Zhou -
2018 Oral: Inter and Intra Topic Structure Learning with Word Embeddings »
He Zhao · Lan Du · Wray Buntine · Mingyuan Zhou -
2018 Poster: Semi-Implicit Variational Inference »
Mingzhang Yin · Mingyuan Zhou -
2018 Oral: Semi-Implicit Variational Inference »
Mingzhang Yin · Mingyuan Zhou -
2017 Poster: Deep Latent Dirichlet Allocation with Topic-Layer-Adaptive Stochastic Gradient Riemannian MCMC »
Yulai Cong · Bo Chen · Hongwei Liu · Mingyuan Zhou -
2017 Talk: Deep Latent Dirichlet Allocation with Topic-Layer-Adaptive Stochastic Gradient Riemannian MCMC »
Yulai Cong · Bo Chen · Hongwei Liu · Mingyuan Zhou