Feature Discovery by Competive Learning

By David E. Rumelhart, David Zipser. 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

Deep learning in neural networks: An overview (2014, background): “Many papers of the previous millennium, however, were about unsupervised learning (UL) without a teacher (e.g., Hebb, 1949; von der Malsburg, 1973; Kohonen, 1972, 1982, 1988; Willshaw and von der Malsburg, 1976; Grossberg, 1976a,b; Watanabe, 1985; Rumelhart and Zipser, 1986; Pearlmutter and Hinton, ”
Sum-product networks: A new deep architecture (2011, background): “SPNs can also be viewed as a probabilistic version of competitive learning [27] and sigma-p i networks [25].”
Online Learning with Random Representations (1993, methodology): “Competitive learning methods (Rumelhart and Zipser, 1986; Kohonen, 1990) have been widely used for this purpose.”
Training Deep Spiking Neural Networks Using Backpropagation (2016, methodology): “Using Backpropagation in SNNs In order to derive and apply the backpropagation equations for training SNNs, after summarizing the classical backpropagation method (Rumelhart and Zipser, 1985) we derive differentiable transfer functions for spiking neurons in WTA configuration.”
Word learning emerges from the interaction of online referent selection and slow associative learning (2012, background): “In connectionist models connection weights start from small random values—a necessity in unsupervised learning (e.g., Rumelhart & Zipser, 1986).”
Learning Algorithms for Networks with Internal and External Feedback (1990, background): “Learning (heavily employed in work on unsupervised learning (Kohonen, 1988) (Grossberg, 1976)) may be interpreted as`shifting weight substance' from certain incoming connections of a winner-take-all-unit to other incoming connections (Rumelhart and Zipser, 1986).”
Learning spatio-temporal patterns for predicting object behaviour (2000, background): “The two networks are labelled according to an anology with natural-language processing: the symbol network categorises the object shape and locations at any time (in other words, the characters - or symbols - observed), and the context network categorises the order in which they appear (the context…”
Frequency-sensitive competitive learning for scalable balanced clustering on high-dimensional hyperspheres (2004, methodology): “Even after stemming is carried out and both very rare, as well as very common, terms have been discarded, the residual vocabulary often contains thousands of terms.”
Learning the Distribution of Object Trajectories for Event Recognition (1995, methodology): “We implement the technique using a competitive learning neural network which is taught in an unsupervised manner (see eg.[5, 6]).”
Modeling hippocampal and neocortical contributions to recognition memory: a complementary-learning-systems approach. (2003, unknown): “(For more information on these mechanisms, see also Kanerva, 1988; Kohonen , 1977; Minai & Levy, 1994; Oja, 1982; Rumelhart & Zipser, 1986.)”
A local learning algorithm for dynamic feedforward and recurrent networks (1990, background): “There is an analogy between the NBB and competitive learning (Grossberg, 1976)(Kohonen, 1988)(Rumelhart and Zipser, 1986).”
Metaheuristic design of feedforward neural networks: A review of two decades of research (2017, background): “Contrary to the supervised learning paradigm, there are two other learning forms for the spacial cases of FNNs: 1) the unsupervised learning—for the unlabeled training data [35], and 2) the reinforcement learning—for the training data with insufficient input–output relations [36].”
Information Theoretic Clustering (2002, background): “Competitive networks [18], [31], [46], [39], [21] can be used in clustering procedures [12], [20] where they form Voronoi cells [16], [14] with a minimum variance flavor.”
Feature Extraction Through LOCOCODE (1999, background): “Other approaches favor local codes (e.g., Rumelhart & Zipser, 1986; Barrow, 1987; Kohonen, 1988).”