Mohammad Nasir, Muhammad Azim
(2020)
Outliers And Structural Breaks Detection In Autoregressive Model By Indicator Saturation Approach.
Masters thesis, Universiti Sains Malaysia.
Abstract
The indicator saturation approach is one of the latest methods in the literature that
Can detect both the outlier and structural break dates simultaneously in a financial time series data. As the approach applied general-to-specific modelling in identifying the most significant indicators, gets package in r and autometrics in oxmetrics can handle the concerns of more variables than observations number, t . As far as we are aware of, all the leading researches use autometrics in their research and most of them carried out simple static data generating process, (dgp) in monte carlo simulations to investigate the performance of indicator saturation.
Actions (login required)
 |
View Item |
