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
Proceedings ArticleDOI
DisenCDR: Learning Disentangled Representations for Cross-Domain Recommendation
TL;DR: This work considers a key challenge of CDR: How do the authors transfer shared information across domains, and proposes DisenCDR, a novel model to disentangle the domain-shared and domain-specific information and proposes two mutual-information-based disentanglement regularizers.
Posted Content
Learning from the Best: Rationalizing Prediction by Adversarial Information Calibration
TL;DR: This work trains two models jointly: one is a typical neural model that solves the task at hand in an accurate but black-box manner, and the other is a selector-predictor model that additionally produces a rationale for its prediction.
Posted Content
An Information Theoretic Interpretation to Deep Neural Networks
TL;DR: This paper shows that the features extracted by DNN coincide with the result of an optimization problem, which it is called the "universal feature selection" problem, in a local analysis regime, which has direct operational meaning in terms of the performance for inference tasks.
Proceedings Article
InfoBERT: Improving Robustness of Language Models from An Information Theoretic Perspective
TL;DR: InfoBERT as discussed by the authors proposes two mutual-information-based regularizers for model training: (i) an Information Bottleneck regularizer, which suppresses noisy mutual information between the input and the feature representation; and (ii) a Robust Feature regularizer.
Journal ArticleDOI
Concept Embedding Models
M. Zarlenga,Pietro Barbiero,Gabriele Ciravegna,Giuseppe Marra,Francesco Giannini,Michelangelo Diligenti,Zohreh Shams,Frédéric Precioso,Stefano Melacci,Adrian Weller,Pietro Liò,Mateja Jamnik +11 more
TL;DR: This work proposes Concept Embedding Models, a novel family of concept bottleneck models which goes beyond the current accuracy-vs-interpretability trade-off by learning interpretable high-dimensional concept representations.
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.