CYBERNETIC PREDICTING DEVICES

By A. Ivakhnenko, V. G. Lapa. 1966

How researchers see this paper

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

In Robotics & Control

Citations in the corpus

Why Should We Add Early Exits to Neural Networks? (2020, background): “In fact, some of the earliest models predating back-propagation were iterative [ 28, 7 ] , and (unsupervised) layerwise strategies were instrumental in the training of some early very deep CNN models [ 10 ] .”
Deep learning in neural networks: An overview (2014, methodology): “The units of GMDH nets may have polynomial activation functions implementing Kolmogorov–Gabor polynomials (more general than other widely used NN activation functions, Section 2).”
On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models (2015, background): “The first deep learning NNs, however, date back to the 1960s [113, 245] (certain more recent developments are covered in a survey [139]).”
One Big Net For Everything (2018, background): “5.6.3] can be used to further compress ONE, possibly shrinking it by pruning neurons and connections, as proposed already in 1965 for deep learning multilayer perceptrons [27, 26, 77].”
Deep Convolutional Neural Networks for Image Classification: A Comprehensive Review (2017, methodology): “The first feedforward multilayered neural networks were trained in 1965 (Ivakhnenko & Lapa, 1966), and although they did not use backpropagation, they were perhaps the first deep learning systems (Schmidhuber, 2015).”
Bidirectional Helmholtz Machines (2015, background): “The earliest deep algorithms that had multiple layers of non-linear features can be traced back to Ivakhnenko and Lapa in 1965 [43].”
Medical Image Analysis using Convolutional Neural Networks: A Review (2017, background): “cations e.g., image classification [6, 67], video classification [68], visual tracking [69], speech recognition [70] and natural language processing [71]. The history of deep learning started in 1965 [72] but has recently seen some major advances. This is due to the availability of improved computat”
AI-Powered Social Bots (2017, background): “The ﬁrst work detailing a learning algorithm for supervised deep feedforward multilayer per-ceptrons was developed by Ivakhnenko and Lapa in 1965 [6].”
Greedy Layerwise Learning Can Scale to ImageNet (2018, background): “It was considered in multiple earlier works Ivakhnenko & Lapa (1965); Fahlman & Lebiere (1990b); Lengellé & Denoeux (1996) on very simple problems and in a climate where deep learning was not a dominant supervised learning approach.”
A comprehensive survey on machine learning for networking: evolution, applications and research opportunities (2018, methodology): “In the meantime, the first algorithm for training MLPNNs with many layers –also known as Deep NN (DNN) in today’s jargon– was published by Ivakhnenko and Lapa in 1965 [210].”
Decoupled Greedy Learning of CNNs (2019, background): “Recently, several authors have revisited the classic (Ivakhnenko & Lapa, 1965; Bengio et al., 2007) approach of supervised greedy layer-wise training of neural networks (Huang et al., 2018; Marquez et al., 2018).”
“Fake News” Is Not Simply False Information: A Concept Explication and Taxonomy of Online Content (2019, background): “One can also use state-of-the-art Deep Learning models that are likely to improve the classification accuracy even more, but they run the risk of not being able to explain “why” it classified a piece of content in a certain category (Ivakhnenko & Lapa, 1973; LeCun et al., 1989).”
Medical image retrieval using deep convolutional neural network (2017, background): “The history of deep learning started in 1965 [35] but has only seen major advances recently with the availability of improved computational capabilities, nonlinearities which allow for deeper network [36,37] and better ways to initialize deep network [38].”
Online Multi-Target Tracking Using Recurrent Neural Networks (2016, unknown): “1966).”
Reinforcement Learning Upside Down: Don't Predict Rewards - Just Map Them to Actions (2019, background): “In that case, C does not have to be an RNN - a multilayer feedforward network (FNN) [21, 52] is sufﬁcient to learn a policy that maps inputs, desired rewards and time horizons to probability distributions over actions.”
Deep Learning in Alzheimer's Disease: Diagnostic Classification and Prognostic Prediction Using Neuroimaging Data (2019, methodology): “d processing units called artificial neurons that were modeled (McCulloch and Pitts, 1943) and developed with the concept of Perceptron (Rosenblatt, 1957;1958) , Group Method of Data Handling (GMDH) (Ivakhnenko and Lapa, 1965;Ivakhnenko, 1968;1971) and the Neocognitron . Efficient error (Fukushima, ”