Pattern recognition using Boltzmann machine
H. Ma · 2002
An approach for pattern recognition using neural networks is proposed. Particularly, a Boltzmann machine, a Hopfield neural net model, is used in pattern recognition with desirable learning ability. The Boltzmann machine features stochastic learning, which acts as the connection dynamics for determining the weights on the connections between the neuron-like cells (processing elements) of different layers in the neural network. An algorithm for pattern recognition using Boltzmann machine is also presented, which could be coded with the C programming language or others to implement the approach for efficient pattern recognition.