Detection of outliers in mixed regressive-spatial autoregressive models |
| |
Authors: | Libin Jin Xiaowen Dai Anqi Shi |
| |
Affiliation: | 1. School of Statistics, Renmin University of China, Beijing, China;2. College of Letters and Science, University of Wisconsin-Madison, Madison, WI, USA |
| |
Abstract: | ABSTRACTThis article studies the outlier detection problem in mixed regressive-spatial autoregressive model. The formulae for testing outliers and their approximate distributions are derived under the mean-shift model and the variance-weight model, respectively. The simulation studies are conducted for examining the power and size of the test, as well as for the detection of outliers when a simulated data contains several outliers. A real data is analyzed to illustrate the proposed method, and modified models based on mean-shift and variance-weight models in which detected outliers are taken into account are suggested to deal with the outliers and confirm theconclusions. |
| |
Keywords: | Mean-shift model Mixed regressive-spatial autoregressive model Outliers Score test Variance-weight model |
|
|