Transfer Learning with Label Noise

By Xiyu Yu, Tongliang Liu, Mingming Gong, Kun Zhang, Kayhan Batmanghelich, Dacheng Tao. 2017

32 citations recorded by the source corpus.

Transfer learning aims to improve learning in target domain by borrowing knowledge from a related but different source domain. To reduce the distribution shift between source and target domains, recent methods have focused on exploring invariant representations that have similar distributions across domains. However, when learning this invariant knowledge, existing methods assume that the labels in source domain are uncontaminated, while in reality, we often have access to source data with noisy labels. In this paper, we first show how label noise adversely affect the learning of invariant representations and the correcting of label shift in various transfer learning scenarios. To reduce the adverse effects, we propose a novel Denoising Conditional Invariant Component (DCIC) framework, which provably ensures (1) extracting invariant representations given examples with noisy labels in source domain and unlabeled examples in target domain; (2) estimating the label distribution in target domain with no bias. Experimental results on both synthetic and real-world data verify the effectiveness of the proposed method.

Subfields: Artificial Intelligence, Cancer Research

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Cancer Research

Citations in the corpus

Conditional variance penalties and domain shift robustness (2017, background): “Causal models thus have the defining advantage that the predictions will be valid even under arbitrarily large interventions on all predictor variables (Haavelmo 1944; Aldrich 1989; Pearl 2009; Schölkopf et al. 2012; Peters et al. 2016; Zhang et al. 2013, 2015; Yu et al. 2017; Rojas-Carulla et al. 2”
Deep learning with noisy labels: exploring techniques and remedies in medical image analysis (2019, background): “On the practical side, many studies have shown the negative impact of noisy labels on the performance of these models in real-world applications [67], [68], [69].”
LTF: A Label Transformation Framework for Correcting Label Shift (2020, methodology): “The existing methods solve covariate shift and target shift using re-weighting methods, which are also used in a wide range of problems, e.g. , label-noise (Liu & Tao, 2015; Yu et al., 2017b; Cheng et al., 2017; Fang et al., 2020) .”