THE EFFECT OF SMOOTHING PARAMETER IN KERNELS AGGREGATION Page No: 4255-4262

By: Siloko, I U1, Ishiekwene, C C2 and Oyegue, F O

Keywords: Kernel density aggregation, kernel density estimator, smoothing parameter, multiplier factor, asymptotic mean integrated squared error (AMISE).

Abstract: Kernel density estimation depends on appropriate smoothing parameter selection in its implementation since the method is mainly for data exploration and visualization purposes. While considering the effect of the smoothing parameter, the form of aggregation employed will determine the size of the smoothing parameter required for better performance. This paper considered two aggregating methods with respect to the asymptotic mean integrated squared error (AMISE) as the criterion function by introducing the multiplier factor that regulates the selection of smoothing parameter in the multiplicative aggregation. The results of the forms of aggregation considered were compared using real life data.



[View Complete Article]