Algorithm 1 Pseudocode for Outlier Removal |
The data to evaluate are imported. |
Statistical parameters (Mean, Covariance) are calculated. |
Calculation of the parameters (New_Data, inverse covariance) to obtain the Mahalanobis distance. |
|
|
Product of New_Data and inverse covariance. |
Critical value is obtained according to a chi-square distribution with the function ncx2inv with MATLAB. |
Data are separated according to the critical value. |
Outliers are removed and maintained matrices are created. |