Breakdown Analysis of Pearson Correlation Coefficient and Robust Correlation Methods
Friday Zinzendoff Okwonu, Bolaji Laro Asaju, Festus Irimisose Arunaye
Abstract
Abstract This paper discussed plug in robust procedure to robustify the Pearson product moment correlation coefficient (PPMCC). The mean of PPMCC is highly susceptible to influential observations hence the PPMCC is not robust against data set that contains substantial amount of influential observations. This study focused on robust plug in techniques with high breakdown points. The performance of these techniques are compared using real and simulated data set. The comparative analysis indicates different degrees of robustness and breakdown based on the percentage of contamination and data modification.
Topics & Concepts
Pearson product-moment correlation coefficientRobustness (evolution)Correlation coefficientCorrelationStatisticsMoment (physics)Data correlationData setSpark plugRobust statisticsComputer scienceMathematicsData miningPhysicsChemistryThermodynamicsClassical mechanicsEstimatorGeometryBiochemistryGeneAdvanced Statistical Methods and ModelsSpectroscopy and Chemometric AnalysesAdvanced Statistical Process Monitoring