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
Noisy Concurrent Training for Efficient Learning under Label Noise
TL;DR: In this article, the authors propose to use the consensus between two models as an additional source of supervision to discourage memorization and over-generalization in deep neural networks, and they demonstrate the effectiveness of their approach on both synthetic and real world noisy benchmark datasets.
Learning, Moving, And Predicting With Global Motion Representations
TL;DR: This dissertation presents methods to robustly estimate camera motion from noisy optical flow estimates by exploiting the global, statistical relationship between the optical flow field and camera motion under projective geometry, and develops methods to learn representations of visual motion directly from unlabeled image sequences.
Posted Content
Learning Independently-Obtainable Reward Functions
Christopher Grimm,Satinder Singh +1 more
TL;DR: A novel method for learning a set of disentangled reward functions that sum to the original environment reward and are constrained to be independently obtainable, and illustrates that this method can learn meaningful reward decompositions in a variety of domains.
Posted Content
Discrete InfoMax Codes for Meta-Learning
TL;DR: This paper presents a method for learning a stochastic encoder that yields discrete p-way codes of length d by maximizing the mutual information between representations and labels, and shows that using shorter codes reduces overfitting in the context of few-shot classification.
Posted Content
Abstraction Mechanisms Predict Generalization in Deep Neural Networks
Alex Gain,Hava T. Siegelmann +1 more
TL;DR: The Cognitive Neural Activation metric (CNA) for DNNs is defined, which is the correlation between information complexity (entropy) of given input and the concentration of higher activation values in deeper layers of the network.
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.