Translated Learning: Transfer Learning across Different Feature Spaces

By Wenyuan Dai, Yuqiang Chen, Gui-Rong Xue, Qiang Yang, Yong Yu. 2008

319 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

Learning to cluster in order to Transfer across domains and tasks (2017, background): “Cross-task Transfer Learning: Features learned when trained for ImageNet classiﬁcation (Rus-sakovsky et al., 2015) have boosted the performance of a variety of vision tasks in a supervised setting.”
Efficient Learning of Domain-invariant Image Representations (2013, methodology): “Recently proposed feature adaptation methods [1, 2, 12, 13, 14, 15] offer a solution by learning a category-independent feature transform that maps target features into the source, pooling all training labels across categories.”
Learning with Augmented Features for Heterogeneous Domain Adaptation (2012, background): “Thus, they cannot deal with the problem where the dimensions of data from the source and target domains are different, which is known as heterogeneous domain adaptation (HDA) (Dai et al., 2009; Yang et al., 2009).”
Transfer learning for speech and language processing (2015, methodology): “An example is the work proposed in [25], where the authors used co-occurrence data to estimate the corresponden ce”
Transferable Contextual Bandit for Cross-Domain Recommendation (2018, methodology): “To bridge incommensurable domains, existing works fall into two categories: TLRisk (Dai et al. 2008) learns a translator from the co-occurrence data to translate features between domains; HTLIC (Zhu et al.”
What you saw is not what you get: Domain adaptation using asymmetric kernel transforms (2011, background): “Other authors have proposed “translating” features between domains; [13] translated features between camera views to transfer activity models, while [8] translated user preferences between text and image domains.”
Methodologies for Cross-Domain Data Fusion: An Overview (2015, background): “There are two directions in this stream of work: 1) translation from the source to the target [18] or 2) projection both domains into a common latent space [72][93].”
Toward Scalable Systems for Big Data Analytics: A Technology Tutorial (2014, background): “Because social networks contain a large amount of linked information among different types of objects, such as articles, tags, images, and videos, transfer learning in heterogeneous networks aims to transfer information knowledge across links [243].”
Hybrid Heterogeneous Transfer Learning through Deep Learning (2014, methodology): “Moreover, there are many methods (Dai et al. 2008; Wang and Mahadevan 2009) proposed to learn the mappings by incorporating instance correspondences between domains.”
Transfer Neural Trees for Heterogeneous Domain Adaptation (2016, background): “, cross-domain object recognition [2–5] and crosslanguage text categorization [6, 7]) can be viewed as DA problems.”
Deep Asymmetric Transfer Network for Unbalanced Domain Adaptation (2018, methodology): “Most of existing methods (Zhu et al. 2011; Qi, Aggarwal, and Huang 2011; Shi et al. 2009; Dai et al. 2009) adopt shallow models attempting to explicitly reduce the discrepancy.”
Big Data: A Survey (2014, background): “Considering that SNS contains massive information, transfer learning in heterogeneous networks aims to transfer knowledge information among different media [150].”