Generalized DataWeighting via Class-Level Gradient Manipulation

Authors: Can Chen, Shuhao Zheng, Xi Chen, Erqun Dong, Xue (Steve) Liu, Hao Liu, Dejing Dou

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments in various settings verify the effectiveness of GDW.
Researcher Affiliation Collaboration 1Mc Gill University, 2The Hong Kong University of Science and Technology, 3Baidu Research
Pseudocode Yes Algorithm 1 Generalized Data Weighting via Class-Level Gradients Manipulation
Open Source Code Yes Our code is available at https://github.com/GGchen1997/GDW-NIPS2021.
Open Datasets Yes We randomly select 100 clean images per class from CIFAR10 [47] as the meta set (1000 images in total). Similarly, we select a total of 1000 images from CIFAR100 as its meta set. ... Long-Tailed CIFAR [47] are created by reducing the number of training instances per class...
Dataset Splits Yes In most classification tasks, there is a training set Dtrain = {(xi, yi)}N i=1 and we assume there is also a clean unbiased meta set Dmeta = {(xv i , yv i )}M i=1. We aim to alleviate label noise and class imbalance in Dtrain with Dmeta. ... We randomly select 100 clean images per class from CIFAR10 [47] as the meta set (1000 images in total).
Hardware Specification No The paper does not provide specific details about the hardware used for running the experiments (e.g., GPU/CPU models, memory).
Software Dependencies No The paper mentions using specific models like Res Net-32 and Res Net-18, but does not provide version numbers for any software dependencies or libraries (e.g., Python, PyTorch, TensorFlow).
Experiment Setup Yes Most of our training settings follow [23] and we use the cosine learning rate decay schedule [49] for a total of 80 epochs for all methods. See Appendix C for details. ... All methods are trained for 5 epochs via SGD with a 0.9 momentum, a 10 3 initial learning rate, a 10 3 weight decay, and a 128 batchsize. See Appendix E for details.