Open AccessSelf-organized criticality of molecular biology and thermodynamic analysis of life system based on optimized particle swarm algorithm
Author(s) -
Jin Li,
Fang Xie
Publication year - 2020
Publication title -
cellular and molecular biology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.371
H-Index - 71
eISSN - 1165-158X
pISSN - 0145-5680
DOI - 10.14715/cmb/2020.66.2.29
Subject(s) - criticality , particle swarm optimization , computer science , thermodynamic system , artificial neural network , principal component analysis , self organized criticality , artificial intelligence , statistical physics , machine learning , physics , thermodynamics , nuclear physics
In order to improve the thermodynamic analysis and prediction ability of biological self-organized criticality and life system, a prediction model of biological self-organized criticality and thermodynamic characteristics of life system based on particle swarm optimization neural network is proposed. Fuzzy regression parameter fusion model is adopted to rearrange the statistical prior data of biological self-organized criticality and thermodynamic characteristics of life system, neural network training method is adopted to extract principal component characteristics of rearranged biological self-organized criticality and thermodynamic information flow of life system, and optimized particle swarm algorithm is adopted to carry out feature selection and self-organized supervised learning on extracted principal component characteristics, thus realizing accurate prediction of biological self-organized criticality and thermodynamic characteristics of life system. The simulation results show that the prediction accuracy of biological self-organization criticality and thermodynamic characteristics of life system using this model is high, the prior sample knowledge required is relatively small, and the reliability of biological self-organization criticality characteristics analysis is guaranteed.