Learning representations of back-propagation errors

By David E. Rumelhart, De Rumerhart, Geoffrey E. Hinton, Ronald J. Williams. 1986

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Relational autoencoder for feature extraction (2017, methodology): “Autoencoder was initially introduced in the later 1980s [33] as a linear feature extraction method.”
Illuminating the “black box”: a randomization approach for understanding variable contributions in artificial neural networks (2002, methodology): “…instead refer the reader to the extensive coverage provided in the texts by Smith (1994) It is sufficient to say that the methods described in this paper refer to the classic family of one hidden-layer, feed-forward neural network trained by the backpropagation algorithm (Rumelhart et al., 1986).”
Handwritten digit recognition: benchmarking of state-of-the-art techniques (2003, methodology): “Neural networks for classi cation include the multilayer perceptron (MLP) [18], the radial basis function (RBF) network [14,19,20], the polynomial classi er [21,22], etc.”
Assessing the Scalability of Biologically-Motivated Deep Learning Algorithms and Architectures (2018, methodology): “Backpropagation [32] was popularized as a method for training neural networks by computing gradients with respect to layer parameters using the chain rule: Thus, gradients are obtained by ﬁrst propagating activations forward to the output layer via eq.”
Lensless computational imaging through deep learning (2017, methodology): “These have included: architectures, such as convolutional connectivity [6–9] for regularization and pruning; nonlinearities, such the now widespread use of non-differentiable piecewise linear units [10] as opposed to the older sigmoidal functions that were differentiable but also prone to stagnation”
Application of Deep Convolutional Neural Networks for Detecting Extreme Weather in Climate Datasets (2016, background): “Most of current deep neural network uses back propagation as learning rule [19].”
Deep Computation Model for Unsupervised Feature Learning on Big Data (2016, methodology): “Moreover, a high-order back-propagation algorithm (HBP) as an extension of the conventional back-propagation algorithm [18] in the high-order tensor space is designed for training the parameters in TAE.”
A hybrid ARIMA and artificial neural networks model to forecast particulate matter in urban areas: The case of Temuco, Chile (2008, background): “The most extensively studied and used ANN models are the multilayer feed forward networks (Rumelhart et al., 1986), which allow information transfer only from an earlier layer to the next consecutive layers.”
BiGGER: A new (soft) docking algorithm for predicting protein interactions (2000, methodology): “Thetraining phase used a neural network back propagation algorithm 30 and was targeted at maximizing the distinction between near-correct and incorrect structures.”
Extreme learning machines: a survey (2011, methodology): “backpropagation (BP) method [1] for multi-layer feedforward neural networks).”
A Species‐Specific Approach to Modeling Biological Communities and Its Potential for Conservation (2003, methodology): “I used a feed-forward neural network trained by the back-propagation algorithm ( Rumelhart et al. 1986 ).”
Extreme Learning Machine for Regression and Multiclass Classification (2012, methodology): “Different from traditional learning algorithms for a neural type of SLFNs [23], ELM aims to reach not only the smallest training error but also the smallest norm of output weights.”
Neuroevolution: from architectures to learning (2008, background): “For example, gradient-based learning algorithms such as back-propagation [68, 69] are sensitive to the initial weight values, which may Fig.”