Transfer Learning via Dimensionality Reduction

By Sinno Jialin Pan, James T. Kwok, Qiang Yang. 2008

724 citations recorded by the source corpus.

How researchers see this paper

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

Citations in the corpus

A Comprehensive Survey on Transfer Learning (2019, neutral): “1) μ = 1 and λ1 = 0: The objective function of MMD embedding (MMDE) is given by [75]”
Transfer Joint Matching for Unsupervised Domain Adaptation (2014, background): “Unfortunately, it is nontrivial to reweight source instances if we also require to match the feature distributions in the inﬁnite-dimensional RKHS. Recent works have performed instance reweighting via kernel mean matching [11, 5], sample selection [28], and co-training [4].”
Structure-Preserved Multi-source Domain Adaptation (2016, methodology): “Typical methods include finding intermediate subspaces that describe the smooth transition from one domain to another [8], [10], [14], [23], learning common features subspace [3], [24], [27], [29], [31] or deep features [6], [28].”
Deep Domain-Adversarial Image Generation for Domain Generalisation (2020, background): “Among the recent alignment based deep DG models, (Li et al. 2018b) proposed to minimise MMD of all possible pairs within source domains, meanwhile an adversarial au-toencoder was used to ensure that the learned features follow the Laplace distribution.”
Unsupervised Visual Domain Adaptation Using Subspace Alignment (2013, methodology): “PCA based DA methods have then been naturally investigated [6, 12, 13] in order to find a common latent space where the difference between the marginal distributions of the two domains is minimized with respect to the Maximum Mean Discrepancy (MMD) divergence.”
Transferrable Prototypical Networks for Unsupervised Domain Adaptation (2019, background): “Based on the widely adopted assumption of the existence of a shared feature space for source and target domains in [16, 20, 29], the ultimate goal of this task is to design an embedding function f (xi; θ) which formally reduces domain shifts in the shared feature space and enables learning of both t”
Transfer learning using computational intelligence: A survey (2015, background): “The feature representation approaches can be categorized into two classes [11]: (A) Distribution similarity approaches aim explicitly to make the source and target domain sample distributions similar, either by penalizing or removing features whose statistics vary between domains [24, 30-32] or by l”
Cross-Domain Recommender Systems (2015, unknown): “…al. 2009 [40] [41] Moreno et al. 2012 [46] Shi et al. 2011 [59] Pan et al. 2012 [52] Gao et al. 2013 [26] Pan et al. 2013 [53] Zhao et al. 2013 [69] Pan et al. 2008 [48] Stewart et al. 2009 [60] Pan et al. 2010 [51] Tiroshi et al. 2013 [64] Loni et al. 2014 [44] Diversity Winoto et al. 2008 [66]…”
A Practical Transfer Learning Algorithm for Face Verification (2013, background): “The KL divergence, as well as alternative penalties based on Bregman divergences and maximum mean discrepancy, have been motivated for related transfer learning purposes [7, 23, 16, 20, 24, 9], although not in combination with a likelihood function as we have done here.”
Learning the Shared Subspace for Multi-task Clustering and Transductive Transfer Classification (2009, background): “Transfer learning can be categorized as (1) inductive transfer: there are a few labeled data in the target task, while there are a large amount of labeled [18] [19] [20] or unlabeled [21] data in the source task, (2) transductive transfer: there are no labeled data in the target task, while there ar”
Transfer Learning with Graph Co-Regularization (2012, methodology): “Distribution Matching: Well-known methods are Maximum Mean Discrepancy Embedding (MMDE) [27], Transfer Component Analysis (TCA) [15], and Transfer Subspace Learning (TSL) [28].”
Adaptation Regularization: A General Framework for Transfer Learning (2014, methodology): “Typical learning strategies includes: 1) Correspondence Learning, which first identifies the correspondence among features and then explores this correspondence for transfer subspace learning [4], [5]; 2) Property Preservation, which extracts shared latent factors between domains by preserving the i”
Learning Compatible Embeddings (2021, background): “[29] propose to learn a low-dimensional latent feature space where the distributions between the source and target domain data are the same or close to each other.”
Semi-supervised Domain Adaptation with Subspace Learning for visual recognition (2015, methodology): “Pan et al. [21] proposed a new dimensionality reduction method called maximum mean discrepancy embedding (MMDE) for domain adaptation, which aims to learn a shared latent space where distance between distributions can be reduced while the data variance can be preserved.”
Unsupervised Domain Adaptive Re-Identification: Theory and Practice (2018, methodology): “In a more general manner, [22] and [4] try to find a feature space in which the source and target distributions are similar and the statistic Maximum Mean Discrepancy (MMD) is employed.”
Stratified Transfer Learning for Cross-domain Activity Recognition (2017, background): “[24] proposed maximum mean discrepancy embedding (MMDE) to learn latent features in the reproducing kernel Hilbert space (RKHS).”