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
Journal ArticleDOI
Probabilistic Ensemble of Deep Information Networks
Giulio Franzese,Monica Visintin +1 more
TL;DR: This work describes a classifier made of an ensemble of decision trees, designed using information theory concepts, that is able to provide results comparable to those of the tree classifier in terms of accuracy, while it shows many advantages interms of modularity, reduced complexity, and memory requirements.
Posted Content
DynamicEmbedding: Extending TensorFlow for Colossal-Scale Applications
Yun Zeng,Siqi Zuo,Dongcai Shen +2 more
TL;DR: A new neuron model, called DynamicCell, is proposed, drawing inspiration from from the free energy principle to introduce the concept of reaction to discharge non-digestive energy, which also subsumes gradient descent based approaches as its special cases.
Journal Article
Measuring Unintended Memorisation of Unique Private Features in Neural Networks
J. Hartley,Sotirios A. Tsaftaris +1 more
TL;DR: It is found that strategies to prevent overfitting (e.g. early stopping, regularisation, batch normalisation) do not prevent memorisation of unique features, implying that neural networks pose a privacy risk to rarely occurring private information.
Proceedings ArticleDOI
RényiCL: Contrastive Representation Learning with Skew Rényi Divergence
Kyungmin Lee,Jinwoo Shin +1 more
TL;DR: This work proposes a novel contrastive objective that conducts variational estimation of a skew Rényi divergence and provides a theoretical guarantee on how variations of skew divergence leads to stable training, and shows that Rényu contrastive learning with stronger augmentations outperforms other self-supervised methods without extra regularization or computational overhead.
Proceedings ArticleDOI
Towards True Lossless Sparse Communication in Multi-Agent Systems
TL;DR: In this paper , an information maximization autoencoder and sparse communication loss are used to reframe sparsity as a representation learning problem, which enables lossless sparse communication at lower budgets than prior art.
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.