Learning Invariant Features Using Subspace Restricted Boltzmann Machine

详细信息    查看全文

• 作者：Jakub M. Tomczak ; Adam Gonczarek
• 关键词：Feature learning ; Unsupervised learning ; Invariant features ; Subspace features ; Deep model
• 刊名：Neural Processing Letters
• 出版年：2017
• 出版时间：February 2017
• 年：2017
• 卷：45
• 期：1
• 页码：173-182
• 全文大小：762KB
• 刊物类别：Physics and Astronomy
• 刊物主题：Artificial Intelligence (incl. Robotics); Complex Systems; Computational Intelligence;
• 出版者：Springer US
• ISSN：1573-773X
• 卷排序：45

文摘

The subspace restricted Boltzmann machine (subspaceRBM) is a third-order Boltzmann machine where multiplicative interactions are between one visible and two hidden units. There are two kinds of hidden units, namely, gate units and subspace units. The subspace units reflect variations of a pattern in data and the gate unit is responsible for activating the subspace units. Additionally, the gate unit can be seen as a pooling feature. We evaluate the behavior of subspaceRBM through experiments with MNIST digit recognition task and Caltech 101 Silhouettes image corpora, measuring cross-entropy reconstruction error and classification error.