More Like this

1. Gandalf: Learning label-label correlations in extreme multi-label classification via label features

   Kharbanda, Siddhant; Gupta, Devaansh; Schultheis, Erik; Banerjee, Atmadeep; Hsieh, Cho-Jui; Babbar, Rohit (August 2024, KDD)
2. Gandalf: Learning Label-label Correlations in Extreme Multi-label Classification via Label Features

   https://doi.org/10.1145/3637528.3672063  

   Kharbanda, Siddhant; Gupta, Devaansh; Schultheis, Erik; Banerjee, Atmadeep; Hsieh, Cho-Jui; Babbar, Rohit (August 2024, ACM)
3. PiCO+: Contrastive Label Disambiguation for Robust Partial Label Learning

   https://doi.org/10.1109/TPAMI.2023.3342650  

   Wang, Haobo; Xiao, Ruixuan; Li, Yixuan; Feng, Lei; Niu, Gang; Chen, Gang; Zhao, Junbo (May 2024, IEEE Transactions on Pattern Analysis and Machine Intelligence)
4. Learning from Label Proportions by Learning with Label Noise

   Zhang, Jianxin; Wang, Yutong; Scott, Clayton (January 2022, Conference on Neural Information Processing Systems (NeurIPS 2022))

   Learning from label proportions (LLP) is a weakly supervised classification problem where data points are grouped into bags, and the label proportions within each bag are observed instead of the instance-level labels. The task is to learn a classifier to predict the labels of future individual instances. Prior work on LLP for multi-class data has yet to develop a theoretically grounded algorithm. In this work, we propose an approach to LLP based on a reduction to learning with label noise, using the forward correction (FC) loss of Patrini et al. [30]. We establish an excess risk bound and generalization error analysis for our approach, while also extending the theory of the FC loss which may be of independent interest. Our approach demonstrates improved empirical performance in deep learning scenarios across multiple datasets and architectures, compared to the leading methods. 

   more » « less
5. Learning from Label Proportions by Learning with Label Noise

   Zhang, Jianxin; Wang, Yutong; Scott, Clayton (January 2022, 36th Conference on Neural Information Processing Systems (NeurIPS 2022))

   Learning from label proportions (LLP) is a weakly supervised classification problem where data points are grouped into bags, and the label proportions within each bag are observed instead of the instance-level labels. The task is to learn a classifier to predict the labels of future individual instances. Prior work on LLP for multi-class data has yet to develop a theoretically grounded algorithm. In this work, we propose an approach to LLP based on a reduction to learning with label noise, using the forward correction (FC) loss of Patrini et al. [30]. We establish an excess risk bound and generalization error analysis for our approach, while also extending the theory of the FC loss which may be of independent interest. Our approach demonstrates improved empirical performance in deep learning scenarios across multiple datasets and architectures, compared to the leading methods. 

   more » « less