Robust regression and outlier detection
Robust regression and outlier detection
Unsupervised Optimal Fuzzy Clustering
IEEE Transactions on Pattern Analysis and Machine Intelligence
Detection of outliers and robust estimation using fuzzy clustering
Computational Statistics & Data Analysis
| Hi-index | 0.00 |
The bootstrap test (BT) is a computer-intensive statistical technique for testing. In practical application, the BT is conducted by using bootstrap samples, and the theoretical distribution of the test statistic is replaced by its empirical distribution under the null hypothesis is assumed. But, it is unreasonable to use the same selected probability in generating bootstrap samples when the data set contains outliers. In this paper, fuzzy weights are used in the BT to reduce the affection of outliers. Every observation is selected according to the fuzzy weights when bootstrap samples are generated. The new BT procedure is called the fuzzy-weighted bootstrap test (FWBT) procedure. Simulation results show that the FWBT is less sensitive to outliers and the test result is stable.