A DNA Computing Model of Perceptron

摘要

Since Adleman shows that DNA strands can be used to solve an instance of the NP-complete Hamiltonian path problem (HPP). DNA computing has been used to solve all kinds of difficult problems including neural networks. This paper put forward a DNA computing model of perceptron to implement linear categorizer function. The advantage of the DNA computing model is the great parallelism of algorithm which will significant improve the efficiency of perceptron categorizer and reduce the running time of perceptron as the same time.