Domain Disentangled Meta-Learning
Domain Disentangled Meta-Learning
复制标题
DOI：
10.1137/1.9781611977653.ch61
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Xin Zhang;Yanhua Li;Ziming Zhang;Zhi-Li Zhang
中科院分区：
文献类型：
--
作者：
Xin Zhang;Yanhua Li;Ziming Zhang;Zhi-Li Zhang
A key challenge with supervised learning (e.g ., image clas-sification) is the shift of data distribution and domain from training to testing datasets, so-called “domain shift” (or “dis-tribution shift”), which usually leads to a reduction of model accuracy. Various meta-learning approaches have been proposed to prevent the accuracy loss by learning an adaptable model with training data, and adapting it to test time data from a new data domain. However, when the域移动发生在多个域维度（例如，图像可以通过旋转，过渡和扩展来转换），适应模型的平均预测能力将确定解决此问题。我们在图像上评估了DDML使用三个数据集的分类问题与多个域的尺寸相比，与元学习和经验风险最小化的各种基础相比，我们的DDML方法始终如一地实现了较高的分类准确性。
A key challenge with supervised learning ( e.g ., image clas-sification) is the shift of data distribution and domain from training to testing datasets, so-called “domain shift” (or “dis-tribution shift”), which usually leads to a reduction of model accuracy. Various meta-learning approaches have been proposed to prevent the accuracy loss by learning an adaptable model with training data, and adapting it to test time data from a new data domain. However, when the domain shift occurs in multiple domain dimensions ( e.g ., images may be transformed by rotations, transitions, and expansions), the average predictive power of the adapted model will deteriorate. To tackle this problem, we propose a domain disentangled meta-learning (DDML) framework. DDML disentangles the data domain by dimensions, learns the representations of domain dimensions independently, and adapts to the domain of test time data. We evaluate our DDML on image classifi-cation problems using three datasets with distribution shifts over multiple domain dimensions. Comparing to various base-lines in meta-learning and empirical risk minimization, our DDML approach achieves consistently higher classification accuracy with the test time data. These results demonstrate that domain disentanglement reduces the complexity of the model adaptation, thus increases the model generalizability, and prevents it from overfitting.