Outlier Detection of Functional Data Using Reproducing Kernel Hilbert Space

The problem of finding the pattern that deviates from other observation is termed as outlier. detection outlier getting importance in research area nowadays due to reason technique has been used various mission critical applications such military, health care, fault recovery, and many. analysis functional data its depth function plays a crucial role statistical model for detecting values alone not enough outliers, since all low be an main using classical it cannot cop up with high dimensionality This paper proposed novel based on Reproducing Kernel Hilbert Space curve (RKHS) outliers data. RKHS special space associated kernel so reproduces each enhance performance function. method uses distance weighted discrimination classification avoids overfitting provides better generalizability dimensions. depths perform performances number artificial real sets.