Opening the Black Box: predicting the trainability of deep neural networks with reconstruction entropy

• URL: http://arxiv.org/abs/2406.12916v3
• Date: Thu, 19 Dec 2024 19:33:28 GMT
• Title: Opening the Black Box: predicting the trainability of deep neural networks with reconstruction entropy
• Authors: Yanick Thurn, Ro Jefferson, Johanna Erdmenger,
• Abstract summary: We present a method for predicting the trainable regime in parameter space for deep feedforward neural networks (DNNs) We show that a single epoch of training is sufficient to predict the trainability of the deep feedforward network on a range of datasets.
• Score: 0.0
• License:
• Abstract: An important challenge in machine learning is to predict the initial conditions under which a given neural network will be trainable. We present a method for predicting the trainable regime in parameter space for deep feedforward neural networks (DNNs) based on reconstructing the input from subsequent activation layers via a cascade of single-layer auxiliary networks. We show that a single epoch of training of the shallow cascade networks is sufficient to predict the trainability of the deep feedforward network on a range of datasets (MNIST, CIFAR10, FashionMNIST, and white noise), thereby providing a significant reduction in overall training time. We achieve this by computing the relative entropy between reconstructed images and the original inputs, and show that this probe of information loss is sensitive to the phase behaviour of the network. We further demonstrate that this method generalizes to residual neural networks (ResNets) and convolutional neural networks (CNNs). Moreover, our method illustrates the network's decision making process by displaying the changes performed on the input data at each layer, which we demonstrate for both a DNN trained on MNIST and the vgg16 CNN trained on the ImageNet dataset. Our results provide a technique for significantly accelerating the training of large neural networks.

Related papers

• Deep Neural Networks Tend To Extrapolate Predictably [51.303814412294514]
  neural network predictions tend to be unpredictable and overconfident when faced with out-of-distribution (OOD) inputs. We observe that neural network predictions often tend towards a constant value as input data becomes increasingly OOD. We show how one can leverage our insights in practice to enable risk-sensitive decision-making in the presence of OOD inputs.
  arXiv  Detail & Related papers  (2023-10-02T03:25:32Z)
• Neural networks trained with SGD learn distributions of increasing complexity [78.30235086565388]
  We show that neural networks trained using gradient descent initially classify their inputs using lower-order input statistics. We then exploit higher-order statistics only later during training. We discuss the relation of DSB to other simplicity biases and consider its implications for the principle of universality in learning.
  arXiv  Detail & Related papers  (2022-11-21T15:27:22Z)
• Reconstructing Training Data from Trained Neural Networks [42.60217236418818]
  We show in some cases a significant fraction of the training data can in fact be reconstructed from the parameters of a trained neural network classifier. We propose a novel reconstruction scheme that stems from recent theoretical results about the implicit bias in training neural networks with gradient-based methods.
  arXiv  Detail & Related papers  (2022-06-15T18:35:16Z)
• Data-driven emergence of convolutional structure in neural networks [83.4920717252233]
  We show how fully-connected neural networks solving a discrimination task can learn a convolutional structure directly from their inputs. By carefully designing data models, we show that the emergence of this pattern is triggered by the non-Gaussian, higher-order local structure of the inputs.
  arXiv  Detail & Related papers  (2022-02-01T17:11:13Z)
• Training Graph Neural Networks by Graphon Estimation [2.5997274006052544]
  We propose to train a graph neural network via resampling from a graphon estimate obtained from the underlying network data. We show that our approach is competitive with and in many cases outperform the other over-smoothing reducing GNN training methods.
  arXiv  Detail & Related papers  (2021-09-04T19:21:48Z)
• Feature Alignment for Approximated Reversibility in Neural Networks [0.0]
  We introduce feature alignment, a technique for obtaining approximate reversibility in artificial neural networks. We show that the technique can be modified for training neural networks locally, saving computational memory resources.
  arXiv  Detail & Related papers  (2021-06-23T17:42:47Z)
• Local Critic Training for Model-Parallel Learning of Deep Neural Networks [94.69202357137452]
  We propose a novel model-parallel learning method, called local critic training. We show that the proposed approach successfully decouples the update process of the layer groups for both convolutional neural networks (CNNs) and recurrent neural networks (RNNs) We also show that trained networks by the proposed method can be used for structural optimization.
  arXiv  Detail & Related papers  (2021-02-03T09:30:45Z)
• Implicit recurrent networks: A novel approach to stationary input processing with recurrent neural networks in deep learning [0.0]
  In this work, we introduce and test a novel implementation of recurrent neural networks into deep learning. We provide an algorithm which implements the backpropagation algorithm on a implicit implementation of recurrent networks. A single-layer implicit recurrent network is able to solve the XOR problem, while a feed-forward network with monotonically increasing activation function fails at this task.
  arXiv  Detail & Related papers  (2020-10-20T18:55:32Z)
• Modeling from Features: a Mean-field Framework for Over-parameterized Deep Neural Networks [54.27962244835622]
  This paper proposes a new mean-field framework for over- parameterized deep neural networks (DNNs) In this framework, a DNN is represented by probability measures and functions over its features in the continuous limit. We illustrate the framework via the standard DNN and the Residual Network (Res-Net) architectures.
  arXiv  Detail & Related papers  (2020-07-03T01:37:16Z)
• Backprojection for Training Feedforward Neural Networks in the Input and Feature Spaces [12.323996999894002]
  We propose a new algorithm for training feedforward neural networks which is fairly faster than backpropagation. The proposed algorithm can be used for both input and feature spaces, named as backprojection and kernel backprojection, respectively.
  arXiv  Detail & Related papers  (2020-04-05T20:53:11Z)
• Curriculum By Smoothing [52.08553521577014]
  Convolutional Neural Networks (CNNs) have shown impressive performance in computer vision tasks such as image classification, detection, and segmentation. We propose an elegant curriculum based scheme that smoothes the feature embedding of a CNN using anti-aliasing or low-pass filters. As the amount of information in the feature maps increases during training, the network is able to progressively learn better representations of the data.
  arXiv  Detail & Related papers  (2020-03-03T07:27:44Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.