Expectation maximization transfer learning and its application for bionic hand prostheses
Expectation maximization transfer learning and its application for bionic hand prostheses
复制标题
DOI：
10.1016/j.neucom.2017.11.072
复制
发表时间：
2018-07-12
期刊：
影响因子：
6
通讯作者：
Hammer, Barbara
中科院分区：
文献类型：
--
作者：
Paassen, Benjamin;Schulz, Alexander;Hammer, Barbara
实际环境中的机器学习模型通常面临输入数据分布的变化。这样的改变会严重影响模型的性能，例如导致数据的错误分类。这一点在仿生假手领域尤为明显，机器学习模型承诺提供更快、更直观的用户界面，但由于缺乏对日常干扰(如电极移位)的稳健性而受到阻碍。解决数据分布变化的一种方法是转移学习，即将受干扰的数据转移到原始模型再次适用的空间。在这一贡献中，我们提出了一种新的期望最大化算法来学习线性变换，根据未扰动模型最大化扰动数据的可能性。我们还证明了这种方法可以推广到判别模型，特别是学习矢量量化模型。在我们对仿生假肢领域的数据的评估中，我们证明了我们的方法可以学习一种变换，如果目标领域中有很少的数据或几个类别可用，那么我们的方法可以显著提高分类精度，并且性能优于所有测试的基线。(C)2018爱思唯尔B.V.保留所有权利。
Machine learning models in practical settings are typically confronted with changes to the distribution of the incoming data. Such changes can severely affect the model performance, leading for example to misclassifications of data. This is particularly apparent in the domain of bionic hand prostheses, where machine learning models promise faster and more intuitive user interfaces, but are hindered by their lack of robustness to everyday disturbances, such as electrode shifts. One way to address changes in the data distribution is transfer learning, that is, to transfer the disturbed data to a space where the original model is applicable again. In this contribution, we propose a novel expectation maximization algorithm to learn linear transformations that maximize the likelihood of disturbed data according to the undisturbed model. We also show that this approach generalizes to discriminative models, in particular learning vector quantization models. In our evaluation on data from the bionic prostheses domain we demonstrate that our approach can learn a transformation which improves classification accuracy significantly and outperforms all tested baselines, if few data or few classes are available in the target domain. (C) 2018 Elsevier B.V. All rights reserved.