P. Chou. The Capacity of the Kanerva Associative Memory is Exponential. Stanford University CA 94305, Stanford, USA.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....distribution can be defined as: P(X (n) k; p) n k p k q n Gammak ; q = 1 Gamma p; k 2 f0; ng 5 This measure must not necessarily define a metric on the f0; 1g n . 6 Arguments to these functions are memory items located in the space f0; 1g n . 7 Compare [8, 4, 1] for deduction. 5 design parameters activation K p one = 0.5 dist = xor k = n p(h; n; 0.5) SC p one = 0.5 dist = xor h = 0 p(0; k; 0.5) I p one = 0.5 dist = xor p(h; k; 0.5) AK p one = 0.1 dist = xor k = n p(h; n; 0.82) HP p one = 0.1 dist = nand h = 0 p(0; k; 0.9) IH p one = 0.1 dist = ....

P. Chou. The Capacity of the Kanerva Associative Memory is Exponential. Stanford University CA 94305, Stanford, USA.

....bit corruption. Let k = N , where is the fraction of incorrect bits in the pattern S, and define 0 as 1 N . In [3] it was shown that when 0 (1 a 2 ) Gamma1 the asymptotic storage capacity of the ECAM is the largest possible for an associative memory, as determined in [9]. Observe that, under our memoryless corruption assumption, the bit error probability p represents the expected fraction of incorrect bits in any input pattern, and is therefore intimately related to the parameter defined above. Specifically, is the actual realisation of a random variable whose ....

P. A. Chou, "The capacity of the Kanerva associative memory is exponential," in D. Z. Anderson, Ed., Neural Information Processing Systems. New York, NY: American Institute of Physics, 1988, pp. 184--191.

....(ch. 8) His solution assembles separate SDMs into a folded SDM of a definite depth performing an accurate indexing on those sequences. A k folded SDM employs k inner SDMs, each one acting on individually delayed addresses. Sec. 5 gives an example. 2 Compare [Kanerva, 1992; Jaeckel, 1989a; Chou, 1988] for a deduction. 3 Algorithms and Data Types The CM 2 hardware architecture is almost optimally well suited to the requirements of SDM calculations. A large number of processors can be connected efficiently in the form of a virtual grid of arbitrary dimensions. Each processor is equipped with ....

P.A. Chou. The Capacity of the Kanerva Associative Memory is Exponential. Stanford University CA 94305, Stanford, USA, 1988.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC