Your Cart

Your cart is empty

Home  >  Volume 46 (May, 2018)

12. EVALUATION OF PARAMETRIC AND NONPARAMETRIC-BASED CHANGE-POINT CONTROL CHARTS by Salisu S.U., Edokpa W. I., Elakhe O.S. and Shaib I.O. Volume 46 (May, 2018 Issue), pp95 –100
Sale price: $5.00
Quantity:

EVALUATION OF PARAMETRIC AND NONPARAMETRIC-BASED CHANGE-POINT CONTROL CHARTS

Salisu S.U., Edokpa W. I., Elakhe O.S. and Shaib I.O.

Department of Statistics, Federal Polytechnic, Auchi, 

Department of Mathematics, Ambrose Alli University, Ekpoma, Nigeria.

Abstract

When a shift occurs in a process one expects to receive a signal from a control chart as quickly as possible. Upon the receipt of such signal from the control chart a search for the identification for the source of error will commence. The inability of the traditional control charts to indicate the time at which this variation(s) occur poses problems to both manufacturers and engineers. Hence the need to identify the period of change/shift becomes inevitable. The study compares performance of the parametric and nonparametric change-point control charts at different varying start-up using real life data. The concept of maximum likelihood ratio change-point (MLRCP) (parametric-base) performs lower than the Mann-Whitney rank sum test change-point (MWCP) (nonparametric-base) and much lower than the Lapage-change-point (LCP) which is the combination of both Man-Whitney and the Mood Statistics. The LCP does not only detect the shift in location but also the variability. It also reduces the rate of false alarm as the change-point signals shows the actual time of shift. This therefore, shows a clear indication that LCP performs quite better with the competing charts considered. The paper recommends consideration of LPC as a preferred method for the monitoring scheme.

Keywords: Lapage-change-point, Man-Whitney, maximum likelihood ratio and change-point.   

click here to download full Abstract
Close
Loading...