Comparison of Robust Estimators’ Performance for Detecting Outliers in Multivariate Data

In multivariate data, outliers are difficult to detect especially when the dimension of the data increase. Mahalanobis distance (MD) has been one of the classical methods to detect outliers for multivariate data. However, the classical mean and covariance matrix in MD suffered from masking and swamping effects if the data contain outliers. Due to this problem, many studies used a robust estimator instead of the classical estimator of mean and covariance matrix. In this study, the performance of five robust estimators namely Fast Minimum Covariance Determinant (FMCD), Minimum Vector Variance (MVV), Covariance Matrix Equality (CME), Index Set Equality (ISE), and Test on Covariance (TOC) are investigated and compared. FMCD has been widely used and is known as among the best robust estimator. However, there are certain conditions that FMCD still lacks. MVV, CME, ISE and TOC are innovative of FMCD. These four robust estimators improve the last step of the FMCD algorithm. Hence, the objective of this study is to observe the performance of these five estimator to detect outliers in multivariate data particularly TOC as TOC is the latest robust estimator. Simulation studies are conducted for two outlier scenarios with various conditions. There are three performance measures, which are pout, pmask and pswamp used to measure the performance of the robust estimators. It is found that the TOC gives better performance in pswamp for most conditions. TOC gives better results for pout and pmask for certain conditions.

Outlier Detection Using Minimum Vector Variance Algorithm with Depth Function and Mahalanobis Distance

Outliers are observations where the point of observation deviates from the data pattern. The existence of outliers in the data can cause irregularities in the results of data analysis. One solution to this problem is to detect outliers using a statistical approach. The statistical approach method used in this study is the Minimum Vector Variance (MVV) algorithm which has robust characteristics for outliers. The purpose of this research is to detect outliers using the MVV algorithm by changing the data sorting criteria using the Robust Depth Mahalanobis to produce maximum detection. The results obtained from this study are that RDMMVV is superior to the observed value in showing the outliers and the location of the outliers in the data plot compared to DMMVV and MMVV.

Modifikasi Penaksir Robust dalam Pelabelan Outlier Multivariat

Outlier adalah suatu observasi yang polanya tidak mengikuti mayoritas data. Outlier dalam kasus multivariat sangat sulit untuk dideteksi, khususnya ketika dimensi lebih dari 2. Kesulitan ini meningkat ketika data set berukuran besar, yakni jumlah variabel menjadi besar. Metode-metode pendeteksian outlier telah lama berkembang dan beberapa digunakan untuk pelabelan outlier sehingga data dapat dipisahkan antara data yang dicurigai sebagai outlier dan data set pada umumnya. Metode-metode tersebut adalah minimum volume ellipsoid disingkat MVE, minimun covariance determinant disingkat MCD, dan minimum vector variance disingkat MVV. Dari ketiga metode tersebut MVV memiliki waktu perhitungan yang paling cepat. Berdasarkan algoritma MVV, kriteria mengurutkan data menggunakan jarak mahalanobis, maka pada paper ini akan dimodifikasi kriteria pengurutan data dengan menghindari penulisan dalam bentuk invers dari matriks variansi kovariansi. Hasil yang diperoleh adalah metode MVV menjadi lebih cepat dengan menggunakan kriteria baru dengan kecermatan yang sama dengan MVV sebelumnya serta akan diaplikan untuk data real dan data simulasi.