Recognition of printed Chinese numerals based on DNA strand displacement neural network

PDF(2780 KB)
PDF(2780 KB)
Journal of Guangzhou University(Natural Science Edition) ›› 2023, Vol. 22 ›› Issue (1) : 1-8.

Recognition of printed Chinese numerals based on DNA strand displacement neural network

  • WU Qikun, LYU Shuying, SHI Xiaolong
Author information +
History +

Abstract

DNA Strand Displacement ( DSD) is an in vitro thermostatic enzyme free DNA amplification technique, which has become a common technique in the field of DNA computing in recent years. Artificial neural network is a computational model that mimics the structure and function of bio logical neural networks. Based on DNA strand replacement technique, artificial neural networks can be constructed as classifiers for performing various pattern recognition tasks. In this paper, a winner take all ( WTA) neural network is constructed based on DNA strand replacement reaction to accomplish the recognition task of printed Chinese numerals. First, the pictures representing digital patterns are transformed into molecular patterns represented by DNA strands, and then they are fed into the constructed DNA neural network. This network can perform biocomputations using DNA strand dis placement technology to classify input patterns, and the final classification results will be represented by the output of single stranded DNA with fluorescent molecules. Through simulation and biological experiments, we demonstrate that a winner take all neural network based on DNA strand replacement can perform the task of recognition of printed Chinese numerals with excellent performance.

Key words

DNA strand displacement / neural network / pattern recognition

Cite this article

Download Citations
Recognition of printed Chinese numerals based on DNA strand displacement neural network. Journal of Guangzhou University(Natural Science Edition). 2023, 22(1): 1-8

References

[33]MetropolisN,UlamS. The montecarlome thod[J]. Journalof the American Statistical Association,1949,44:33534
PDF(2780 KB)

1117

Accesses

0

Citation

Detail

Sections
Recommended

/