Tensor Regression Networks with various Low-Rank Tensor Approximations
This work addresses the regularization effects of low-rank tensor constraints in neural networks, offering incremental insights for model compression and efficiency in machine learning.
The paper investigates tensor regression networks with various low-rank tensor approximations to compare their compressive and regularization effects on neural networks. Results show that Global Average Pooling outperforms Tensor Regression Layers in deep CNNs on CIFAR-10, while shallow CNNs with tensor regression and dropout achieve lower test errors with limited samples.
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.