Study And Application Of Markov Chain Based On Absorbent-like Wall And Neural Network Model

Both Markov chain model and neural network model are popular mathematical models in recent years.Markov chain model makes use of the state transition law to predict the state of random process,which is largely dependent on previous data.While neural network model obtains results by setting objective function and using big data training,which is less dependent on previous data.This paper studies their properties and characteristics in detail,and tries to combine them to complement each other.The Markov chain model,the neural network model and the Markov neural network model are studied as follows:For the Markov chain model,according to the characteristics of each state in the definition,a new Markov chain model based on the absorbing wall is constructed,and the properties of this model are deduced and studied.The latest time for the state to occur can be calculated by presupposing the probability of occurrence of the agreed state,so as to achieve the effects of early prevention and efficient spot checks.Finally,the model is used to analyze the operating status of the machine that produces men's pants,and the best spot check interval time is obtained.As for the neural network model,the ROC curve based on neural network model is proposed to solve the problem of low classification accuracy caused by the large difference of various scales in big data classification.The objective function of the neural network is changed to the area under the ROC curve,and the AUC value of the ROC curve is used to identify the advantages and disadvantages of the neural network classifier,in order to solve the problem of accurate classification of unbalanced data in the process of big data processing.Finally,the bank default data is analyzed to distinguish the defaulters more effectively,and the practicability of the model is proved.For the Markov neural network model,the validity evaluation model based on Markov chain is constructed to solve the problem that the characteristics of the samples in the time series may change with time passing.The model uses the output results of the neural network to construct a Markov chain,which in turn monitors the timeliness of the neural network.Finally,empirical analysis is made on the stock data of Kweichow Moutai to prove the validity of the model.