Transfer Learning with Graph Co-Regularization

By Mingsheng Long, Jianmin Wang, Guiguang Ding, Dou Shen, Qiang Yang. 2012

204 citations recorded by the source corpus.

Transfer learning is established as an effective technology to leverage rich labeled data from some source domain to build an accurate classifier for the target domain. The basic assumption is that the input domains may share certain knowledge structure, which can be encoded into common latent factors and extracted by preserving important property of original data, e.g., statistical property and geometric structure. In this paper, we show that different properties of input data can be complementary to each other and exploring them simultaneously can make the learning model robust to the domain difference. We propose a general framework, referred to as Graph Co-Regularized Transfer Learning (GTL), where various matrix factorization models can be incorporated. Specifically, GTL aims to extract common latent factors for knowledge transfer by preserving the statistical property across domains, and simultaneously, refine the latent factors to alleviate negative transfer by preserving the geometric structure in each domain. Based on the framework, we propose two novel methods using NMF and NMTF, respectively. Extensive experiments verify that GTL can significantly outperform state-of-the-art learning methods on several public text and image datasets.

Subfields: Artificial Intelligence

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

Citations in the corpus

Heterogeneous Few-Shot Model Rectification With Semantic Mapping (2020, background): “DA aligns distributions between domains and improves the discerning ability of the target classiﬁer [14], [29].”
Deep Model Based Domain Adaptation for Fault Diagnosis (2017, methodology): “One is the CA on the test data, which is widely adopted in the published literature [5], [23], [25], [26].”
Transfer Learning for Semisupervised Collaborative Recommendation (2016, methodology): “Different transfer-learning techniques have been designed for making use of data from different sources in various applications, including text/image mining [Dai et al. 2007; Long et al. 2014] and social computing [Li et al.”
Spectral Bisection Tree Guided Deep Adaptive Exemplar Autoencoder for Unsupervised Domain Adaptation (2016, methodology): “Recently, regularizers such as low-rank constraint (Shao, Kit, and Fu 2014) and Maximum Mean Discrepancy (MMD) (Pan et al. 2011; Baktashmotlagh et al. 2013; Long et al. 2014c) have been applied on transfer learning to guide the subspace or kernel learning.”
Transfer Learning via Learning to Transfer (2018, background): “• Common latent space based transfer learning algorithms: TCA (Pan et al., 2011), ITL (Shi & Sha, 2012), CMF (Long et al., 2014), LSDT (Zhang et al., 2016), STL (Raina et al., 2007), DIP (Baktashmotlagh et al., 2013) and SIE (Baktashmotlagh et al., 2014).”
Scatter Component Analysis: A Unified Framework for Domain Adaptation and Domain Generalization (2015, methodology): “m Mean Discrepancy [41] – referred to as landmarks. The landmarks are then used to construct multiple auxiliary tasks as a basis for composing domain-invariant features. Transfer Joint Matching (TJM) [16] uses a reweighting strategy as a regularizer based on ‘ 2;1-norm structured sparsity on the sou”
Adaptation Regularization: A General Framework for Transfer Learning (2014, background): “…domains by preserving the important properties of the original data, e.g., statistical property [2], [25], geometric structure [26]–[28], or both [3]; 3) Distribution Adaptation, which learns a shared subspace where the distribution difference is explicitly reduced by minimizing predefined…”