Probabilistic Information Capacity of Hopfield Associative Memory
نویسندگان
چکیده
T his pap er defines a form al pr obabilist ic notion for the information capac ity of the Hopfield neur al network model of associat ive memory. A mathematical express ion is derived for the number of random binary pat terns that can be stored as stable states in a Hopfield model of memory with n neur ons with a given probab ility. The derivati on is based on a new approach using two powerfu l mathematical te chniques : Brown 's Mar tingale Central Limit T heorem and Gupta 's tran sformat ion of the pr obab ility integral for a spec ial case of the corr elation matrix. The new approach provides a way for rigorously ana lyzing the complex dynamics of the Hopfield model. Ou r approach refines t he cur rent heurist ic methods, which rely on simplifying ass umptions abo ut the dynamics of the model.
منابع مشابه
Information capacity of the Hopfield model
The information capacity of general forms of memory is formalized. The number of bits of information that can be stored in the Hopfield model of associative memory is estimated. It is found that the asymptotic information capacity of a Hopfield network of N neurons is of the order N3 b. The number of arbitrary state vectors that can be made stable in a Hopfield network of N neurons is proved to...
متن کاملCapacity of oscillatory associative-memory networks with error-free retrieval.
Networks of coupled periodic oscillators (similar to the Kuramoto model) have been proposed as models of associative memory. However, error-free retrieval states of such oscillatory networks are typically unstable, resulting in a near zero capacity. This puts the networks at disadvantage as compared with the classical Hopfield network. Here we propose a simple remedy for this undesirable proper...
متن کاملThe capacity of the Hopfield associative memory
Techniques from,coding theory are applied to study rigorously the capacity of the Hopfield associative memory. Such a memory stores n -tuple of + 1’s. The components change depending on a hardlimited version of linear functions of all other components. With symmetric connections between components, a stable state is ultimately reached. By building up the connection matrix as a sum-of-outer prod...
متن کاملTerminal Attractor Optical Associative Memory for Pattern Recognition
Optical associative memory with terminal attractor (TA) is proposed for pattern recognition. With numerical simulations, the optimal control parameter in the TA model associative memory is determined. The optimal control parameter is also used in an optical experiment. The capacity of TA model associative memory is also investigated based on the consistency between the stored pattern and the ob...
متن کاملCapacities of multiconnected memory models
2014 We study generalizations of the Hopfield model for associative memory which contain interactions of R spins with one another and allow for different weights for input patterns. Using probabilistic considerations we show that stability criteria lead to capacities which increase like powers of NR-1. Investigating numerically the basins of attraction we find behaviour which agrees with theore...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- Complex Systems
دوره 6 شماره
صفحات -
تاریخ انتشار 1992