Calibrate and Debias Layer-wise Sampling for Graph Convolutional Networks

1 code implementation • 1 Jun 2022 • Yifan Chen, Tianning Xu, Dilek Hakkani-Tur, Di Jin, Yun Yang, Ruoqing Zhu

This paper revisits the approach from a matrix approximation perspective, and identifies two issues in the existing layer-wise sampling methods: suboptimal sampling probabilities and estimation biases induced by sampling without replacement.

On Variance Estimation of Random Forests with Infinite-Order U-statistics

1 code implementation • 18 Feb 2022 • Tianning Xu, Ruoqing Zhu, Xiaofeng Shao

To bridge these gaps in the literature, we propose a new view of the Hoeffding decomposition for variance estimation that leads to an unbiased estimator.

Ensemble Learning

Revisiting Layer-wise Sampling in Fast Training for Graph Convolutional Networks

no code implementations • 29 Sep 2021 • Yifan Chen, Tianning Xu, Dilek Hakkani-Tur, Di Jin, Yun Yang, Ruoqing Zhu

To accelerate the training of graph convolutional networks (GCN), many sampling-based methods have been developed for approximating the embedding aggregation.