On the Information Storage Capacity of Local Learning Rules

By Günther Palm. 1992

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): “…1976a, 1976b; Hebb, 1949; Kohonen, 1972, 1982, 1988; Kosko, 1990; Martinetz, Ritter, & Schulten, 1990; Miller, 1994; Mozer, 1991; Oja, 1989; Palm, 1992; Pearlmutter &Hinton, 1986; Ritter & Kohonen, 1989; Rubner & Schulten, 1990; Sanger, 1989; Saund, 1994; von der Malsburg, 1973; Watanabe,…”
Nonlinear ICA through low-complexity autoencoders (1999, background): “Sparseness [13, 14, 15, 16, 17, 18, 19] is not viewed as anpriori good thing, and is not enforced explicitly, but only if the input data indeed is naturally describable by a sparse code.”
Associative Data Storage and Retrieval in Neural Networks (1996, background): “2 will be called storage capacity (discussed in [41]).”
Feature Extraction Through LOCOCODE (1999, unknown): “…encouraging biologically plausible sparse distributed codes (e.g., Field, 1987; Barlow, 1983; Mozer, 1991; Földiák, 1990; Földiák & Young, 1995; Palm, 1992; Zemel & Hinton, 1994; Field, 1994; Saund, 1994; Dayan & Zemel, 1995; Li, 1995; Olshausen & Field, 1996; Zemel, 1993; Hinton &…”
Transitive responding in animals and humans: Exaptation rather than adaptation? (1998, background): “On the other hand, comparably largerbrained animals, which certainly have a larger absolute number and probably a larger proportion of conditionable, plastic synapses, may be able to afford a configural, somewhat redundant, memory consuming representation (Rensch, 1973; Palm, 1992).”