Hysteresis and Robustness of Hopfield Neural Networks
نویسندگان
چکیده
The effect of noise degradation on the Hopfield neural netwerk is s t u d i e d The notion of a hysteresis nefwork is defined. A noisy HephsSi =urd network is subsequently proven to be a hysteresis network. The effect of the hysteresis phentmenon on the robustness of the HoppeM neural network to noise degradation is then investigated. An eptiraal Heefield neural network is defined as the Hopfield neural network which minimizes an upper-bound on the probability of error. The minimal reindicator of a Hopfield neural network is defined. The upper bound on the probability of error of a noisy Hopfield neural network is &rived io terms of the minimal robustness indicator. We h d l y prove that an eptimal HopfieM neural network is obtained when the miniid robustness indicator is maximized.
منابع مشابه
Hopfield Neural Network with Hysteresis for Maximum Cut Problem
A model of neurons with hysteresis (or hysteresis binary neurons) for the Hopfield neural networks is studied. We prove theoretically that the emergent collective properties of the original Hopfield neural networks also are present in the Hopfield neural networks with hysteresis binary neurons. As an example, the networks are also applied to the maximum cut problem and results of computer simul...
متن کاملA New Algorithm Using Hopfield Neural Network with CHN for N-Queens Problem
A model of neurons with CHN (Continuous Hysteresis Neurons) for the Hopfield neural networks is studied. We prove theoretically that the emergent collective properties of the original Hopfield neural networks also are present in the Hopfield neural networks with continuous hysteresis neurons. The network architecture is applied to the N-Queens problem and results of computer simulations are pre...
متن کاملA New Algorithm for Crossbar Switch Problem Using Hopfield Neural Network with Continuous Hysteresis Neurons
In this paper, we propose a continuous hysteresis neurons Hopfield neural network architecture for efficiently solving crossbar switch problems. A Hopfield neural network architecture with continuous hysteresis and its collective computational properties are studied. It is proved theoretically and confirmed by simulating the randomly generated Hopfield neural network with continuous hysteresis ...
متن کاملCrossbar Switch Problem Solver by Hysteresis Neural Networks
In this paper, we propose a continuous hysteresis neurons (CHN) Hopfield neural network architecture for efficiently solving crossbar switch problems. A Hopfield neural network architecture with continuous hysteresis and its collective computational properties are studied. It is proved theoretically and confirmed by simulating the randomly generated Hopfield neural network with CHN. The network...
متن کاملThe hysteretic Hopfield neural network
A new neuron activation function based on a property found in physical systems--hysteresis--is proposed. We incorporate this neuron activation in a fully connected dynamical system to form the hysteretic Hopfield neural network (HHNN). We then present an analog implementation of this architecture and its associated dynamical equation and energy function.We proceed to prove Lyapunov stability fo...
متن کامل