ANALISIS MODEL REGRESI NONPARAMETRIK SIRKULAR-LINEAR BERGANDA
ABSTRACT: Circular data are
data which the value in form of vector is circular data. Statistic analysis
that is used in analyzing circular data is circular statistics analysis. In
regression analysis, if any of predictor or response variables or both are
circular then the regression analysis used is called circular regression analysis.
Observation data in circular statistic which use direction and time units
usually don’t satisfy all of the parametric assumptions, thus making
nonparametric regression as a good solution. Nonparametric regression function
estimation is using epanechnikov kernel estimator for the linier variables and
von Mises kernel estimator for the circular variable. This study showed that
the result of circular analysis by using circular descriptive statistic is
better than common statistic. Multiple circular-linier nonparametric
regressions with Epanechnikov and von Mises kernel estimator didn’t create
estimation model explicitly as parametric regression does, but create
estimation from its observation knots instead.
Penulis: Komang Candra Ivan
Kode Jurnal: jpmatematikadd160182
