Study On Change Point Estimation Of Piecewise Linear Binary Choice Models

In this paper,the model contains change points,the parameters,change points and number of the model are estimated by different smoothing methods in the piecewise linear binary choice model.The estimation results are obtained,and the large-sample properties of the change points are deduced.The first chapter introduces the theory of change points,and introduces the research status,purpose and content of binary choice model.The second chapter introduce the binary choice model with a single change point.Kernel function method is used to smooth the objective function,and maximizes it to obtain the estimation of parameters and change point of the piecewise linear binary choice model,and obtains their asymptotic normality.When is used to estimate parameters by the method,gives initial values to parameters and is obtained by iteration method.The numerical simulation results show that the parameter and change point estimation are valid and robust,which verifies the asymptotic normality theorem in this chapter.The third chapter studies the binary choice model with multiple change points,by linearization technique to piecewise linear binary choice model is transformed into classical binary choice model,the number of unknown for the change points,first assumption that change points number more than real number of change points,the linearization algorithm is used to estimate the number of change points of the model.The number of change points is screened by BIC information criterion to get the number of change points and thelocation estimation.Monte Carlo simulation results show that the estimation is effective.The last chapter summarizes the content of the paper,refines the innovation of the paper,and looks forward to the possible research work.