Application Of Genetic Algorithm, Bootstrap Method And Metropolis-hastings Algorithm To Lattice QCD Numerical Fitting

In this paper,we introduced the genetic algorithm in the field of artificial intelligence,and explored the possibility of applying it to the fitting of Lattice QCD.The bootstrap method is employed for the estimate of standard derivation since the genetic algorithm can not give the error range.A 68% error estimation method is proposed from the new perspective of probability density.The invasive alien species strategy is proposed to improve the computational efficiency of parallel genetic algorithm.The genetic algorithm is also applied to the pre-processing of MetropolisHastings algorithm,which shortens the preheating process of Metropolis-Hastings algorithm.The synthetic data and the actual data are fitted,and compared with the QMBF software which is commonly used in the field of Lattice QCD,and the above methods and ideas are verified.It is proposed that the delay calculation of the fitness can improve the parallel scale of the genetic algorithm and reduce the running time of the algorithm.The effect of deep learning method on improving invasive alien species strategy is put forward.