Deep learning and the information bottleneck principle
Naftali Tishby,Noga Zaslavsky +1 more
- pp 1-5
TLDR
It is argued that both the optimal architecture, number of layers and features/connections at each layer, are related to the bifurcation points of the information bottleneck tradeoff, namely, relevant compression of the input layer with respect to the output layer.Abstract:
Deep Neural Networks (DNNs) are analyzed via the theoretical framework of the information bottleneck (IB) principle. We first show that any DNN can be quantified by the mutual information between the layers and the input and output variables. Using this representation we can calculate the optimal information theoretic limits of the DNN and obtain finite sample generalization bounds. The advantage of getting closer to the theoretical limit is quantifiable both by the generalization bound and by the network's simplicity. We argue that both the optimal architecture, number of layers and features/connections at each layer, are related to the bifurcation points of the information bottleneck tradeoff, namely, relevant compression of the input layer with respect to the output layer. The hierarchical representations at the layered network naturally correspond to the structural phase transitions along the information curve. We believe that this new insight can lead to new optimality bounds and deep learning algorithms.read more
Citations
More filters
Posted Content
Malicious Network Traffic Detection via Deep Learning: An Information Theoretic View.
TL;DR: The results show that since mutual information remains invariant under homeomorphism, only feature engineering methods that alter the entropy of the dataset will change the outcome of the neural network, and suggests that neural networks that can exploit the convolution theorem are equally accurate as standard convolutional neural networks, and can be more computationally efficient.
Journal ArticleDOI
The role of mutual information in variational classifiers
TL;DR: The authors studied the generalization error of classifiers relying on stochastic encodings which are trained on the cross-entropy loss, which is often used in deep learning for classification problems.
Journal ArticleDOI
Hyperbolic Hierarchical Contrastive Hashing
TL;DR: In this article , a hierarchical hyperbolic K-means (HHCH) algorithm was proposed to construct hierarchical semantic structures adaptively. But the proposed method is not suitable for unsupervised hash codes.
Posted Content
Information Competing Process for Learning Diversified Representations
TL;DR: In this article, a new approach, termed Information Competing Process (ICP), is proposed to enrich the information carried by feature representations, which separates a representation into two parts with different mutual information constraints.
Journal ArticleDOI
Variance-Covariance Regularization Improves Representation Learning
TL;DR: Li et al. as mentioned in this paper proposed Variance-Covariance Regularization (VCR), a regularization technique aimed at fostering diversity in the learned network features, which can promote learned representations that exhibit high variance and minimal covariance, thus preventing the network from focusing solely on loss reducing features.
References
More filters
Proceedings Article
ImageNet Classification with Deep Convolutional Neural Networks
TL;DR: The state-of-the-art performance of CNNs was achieved by Deep Convolutional Neural Networks (DCNNs) as discussed by the authors, which consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax.
Book
Elements of information theory
Thomas M. Cover,Joy A. Thomas +1 more
TL;DR: The author examines the role of entropy, inequality, and randomness in the design of codes and the construction of codes in the rapidly changing environment.
Journal ArticleDOI
Reducing the Dimensionality of Data with Neural Networks
TL;DR: In this article, an effective way of initializing the weights that allows deep autoencoder networks to learn low-dimensional codes that work much better than principal components analysis as a tool to reduce the dimensionality of data is described.
Journal ArticleDOI
Representation Learning: A Review and New Perspectives
TL;DR: Recent work in the area of unsupervised feature learning and deep learning is reviewed, covering advances in probabilistic models, autoencoders, manifold learning, and deep networks.
Book
Learning Deep Architectures for AI
TL;DR: The motivations and principles regarding learning algorithms for deep architectures, in particular those exploiting as building blocks unsupervised learning of single-layer modelssuch as Restricted Boltzmann Machines, used to construct deeper models such as Deep Belief Networks are discussed.