期刊
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
卷 30, 期 12, 页码 3759-3773出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TNNLS.2019.2899037
关键词
Manifolds; Measurement; Data models; Kernel; Learning systems; Visualization; Adaptation models; Discrepancy metric; domain adaptation (DA); domain generation; manifold criterion (MC); transfer learning
类别
资金
- National Science Fund of China [61771079, 91420201, 61472187]
- Chongqing Science and Technology Project [cstc2017zdcy-zdzxX0002, cstc2018jcyjAX0250]
- 973 Program [2014CB349303]
- Program for Changjiang Scholars
In many practical transfer learning scenarios, the feature distribution is different across the source and target domains (i.e., nonindependent identical distribution). Maximum mean discrepancy (MMD), as a domain discrepancy metric, has achieved promising performance in unsupervised domain adaptation (DA). We argue that the MMD-based DA methods ignore the data locality structure, which, up to some extent, would cause the negative transfer effect. The locality plays an important role in minimizing the nonlinear local domain discrepancy underlying the marginal distributions. For better exploiting the domain locality, a novel local generative discrepancy metric-based intermediate domain generation learning called Manifold Criterion guided Transfer Learning (MCTL) is proposed in this paper. The merits of the proposed MCTL are fourfold: 1) the concept of manifold criterion (MC) is first proposed as a measure validating the distribution matching across domains, and DA is achieved if the MC is satisfied; 2) the proposed MC can well guide the generation of the intermediate domain sharing similar distribution with the target domain, by minimizing the local domain discrepancy; 3) a global generative discrepancy metric is presented, such that both the global and local discrepancies can be effectively and positively reduced; and 4) a simplified version of MCTL called MCTL-S is presented under a perfect domain generation assumption for more generic learning scenario. Experiments on a number of benchmark visual transfer tasks demonstrate the superiority of the proposed MC guided generative transfer method, by comparing with the other state-of-the-art methods. The source code is available in https://github.com/wangshanshanCQU/MCTL.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据