Title: Estimation of autoregressive models from data with measurement error
Authors: Jessa Lopez - University of the Philippines-Diliman (Philippines) [presenting]
Abstract: When used in modeling, measurement error in the data influences the model structure, and it can even distort the entire data generating process. A method based on spline smoothing and bootstrap is proposed to estimate an autoregressive process to address the issues associated with data contaminated with measurement error. The simulation study indicates that the method estimates the parameters of an autoregressive model better than the usual conditional least squares estimate when the data is measured with error.