Self-Organization and Associative Memory

By Teuvo Kohonen. 1988

Subfields: Representation & Self-Supervised Learning

How researchers see this paper

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

In Representation & Self-Supervised Learning

Citations in the corpus

Experience Grounds Language (2020, background): “The tension of modeling symbols and distributed representations was nicely articulated by Smolensky (1990), and alternative representations (Kohonen, 1984; Hinton et al., 1986; Barlow, 1989) and approaches to structure and composition (Erk and Padó, 2008; Socher et al., 2012) span decades of…”
Multi-column deep neural networks for image classification (2012, neutral): “ed additional heuristics such as unsupervised pre-training [31,26,2,10] or carefully prewired synapses [29,34]. (3) The DNN of this paper (Fig.1a) have 2-dimensional layers of winner-take-all neurons [17,41] with overlapping receptive ﬁelds whose weights are shared [20,1,35,7]. Given some input patt”
Deep learning in neural networks: An overview (2014, neutral): “…& Mitchison, 1989; Barrow, 1987; Deco & Parra, 1997; Field, 1987; Földiák, 1990; Földiák & Young, 1995; Grossberg, 1976a, 1976b; Hebb, 1949; Kohonen, 1972, 1982, 1988; Kosko, 1990; Martinetz, Ritter, & Schulten, 1990; Miller, 1994; Mozer, 1991; Oja, 1989; Palm, 1992; Pearlmutter &Hinton,…”
On the Binding Problem in Artificial Neural Networks (2020, background): “Early work includes Hopfield networks (Hopfield, 1982), Boltzmann machines (Ackley et al., 1985), and associative memory (Kohonen, 1989).”
An Incremental Self-Organizing Architecture for Sensorimotor Learning and Prediction (2017, methodology): “A subclass of neural network models, namely the self-organizing map (SOM) [21], is able to perform local function approximation by partitioning the input space and learning the dynamics of the underlying process in a localized region.”
Transformation Invariance in Pattern Recognition-Tangent Distance and Tangent Propagation (1996, background): “It is straightforward to modify algorithms such as LVQ (Learning Vector quantization) to use a tangent distance.”
Self-Delimiting Neural Networks (2012, background): “The underlying cause of such neighborhood-preserving weight matrices, however, will not be a traditional pre-wired neighborhood-enforcing learning rule [33, 46], but sheer efﬁciency per se .”
Reactivation of Hippocampal Cell Assemblies: Effects of Behavioral State, Experience, and EEG Dynamics (1999, background): “It has been hirther proposed that the changes occurring at the sjniapse are dependent on the coactivity of the pre- and post-synaptic neurons in questions (Hebb, 1949; Marr, 1971; Kohonen, 1988; McNaughton & Morris, 1987).”
DBSCAN: Past, present and future (2014, methodology): “Examples of the model-based clustering algorithms are COBWEB [12] and SOM [13].”
Neural Network Approaches to Image Compression (1995, background): “Recently, there has been an increased interest in the investigation of parallel information processing systems [1], [2], [3], [4], [5].”
Feature Extraction Through LOCOCODE (1999, background): “Other approaches favor local codes (e.g., Rumelhart & Zipser, 1986; Barrow, 1987; Kohonen, 1988 ).”
Learning to Generate Artificial Fovea Trajectories for Target Detection (1991, methodology): “Just like Kohonen's self organizing feature maps [3] dedicate more storage capacity for ne grained representation of common similar inputs, M should 7”
Detection and localization of faces on digital images (1994, methodology): “The general training base is compressed using a Kohonen network (Kohonen (1984)), taking prot of the Vector Quantization properties of this algorithm.”
Fast Online Q(λ) (1998, methodology): “Typicalonline Q(λ) implementations based on lookup tables or other local approximators such as CMACs (Albus, 1975; Sutton, 1996) or self-organizing maps (Kohonen, 1988), however, are unnecessarily time-consuming.”
Kernel Methods and Machine Learning (2014, methodology): “This is exemplified by the well-known SOM clustering algorithm proposed by Kohonen [129, 130, 131].”