Parameter Estimation And Prediction On The Time Series Under Censored Data

In many areas, such as agricultural industry, science and technology, social and eco-nomic life, there are various random phenomena commonly occurring by time series. Time series analysis is to observe,research and extract useful information in order to identify the objective development of things. Statistical time series analysis is the application of the important branch of economy. In many areas such as finance, meteorology and hydrology, we always use time series to build models and predictions.When the data is complete data, we have perfect research on time series. However, censored data is commonly seen in our daily life and its analysis method is more compli-cated than complete data. We need to research the special way to deal with this type of data according to the complete data analysis.Censored data contains many types:left censored, right censored, internal censored. This thesis based on the [1],[2] and [12], makes some research on the parameter estima-tion of time series under censored data. In the end, give some simulations to prove the feasibility of this method.