Current browse context:
cs.LG
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: Tensor Regression Networks with various Low-Rank Tensor Approximations
(Submitted on 27 Dec 2017 (v1), last revised 29 Nov 2018 (this version, v2))
Abstract: Tensor regression networks achieve high compression rate of neural networks while having slight impact on performances. They do so by imposing low tensor rank structure on the weight matrices of fully connected layers. In recent years, tensor regression networks have been investigated from the perspective of their compressive power, however, the regularization effect of enforcing low-rank tensor structure has not been investigated enough. We study tensor regression networks using various low-rank tensor approximations, aiming to compare the compressive and regularization power of different low-rank constraints. We evaluate the compressive and regularization performances of the proposed model with both deep and shallow convolutional neural networks. The outcome of our experiment suggests the superiority of Global Average Pooling Layer over Tensor Regression Layer when applied to deep convolutional neural network with CIFAR-10 dataset. On the contrary, shallow convolutional neural networks with tensor regression layer and dropout achieved lower test error than both Global Average Pooling and fully-connected layer with dropout function when trained with a small number of samples.
Submission history
From: Guillaume Rabusseau [view email][v1] Wed, 27 Dec 2017 08:04:34 GMT (273kb,D)
[v2] Thu, 29 Nov 2018 02:10:55 GMT (1081kb,D)
Link back to: arXiv, form interface, contact.