|
Estimation of
Matusita
Overlapping
Coefficient p
for Pair Normal
Distributions
Omar M. Eidous (1) and Salam K. Daradkeh (2)
(1,2) Department
of Statistics,
Faculty of
Science, Yarmouk
University,
Irbid, Jordan
Email address:
(1)
omarm@yu.edu.jo
Email address:
(2) [Corresponding
Author]
2018107008@ses.yu.edu.jo
Doi :
https://doi.org/10.47013/15.4.23
Cited by :
Jordan J. Math &
Stat.,
15 (4B) (2022),
1137 - 1151
PDF
Received on:
Nov. 12,
2021;
Accepted
on: Feb. 28,
2022
Abstract. The Matusita
overlapping
coefficient ρ is
defined as
agreement or
similarity
between two or
more
distributions.
The parametric
normal
distribution is
one of the most
important
statistical
distributions.
Under the
assumption that
the data at hand
follow two
independent
normal
distributions,
this paper
suggests a new
technique to
estimate the
Matusita
coefficient ρ.
In contrast to
the studies in
the literature,
the suggested
technique
requires no
assumptions on
the location and
scale parameters
of the normal
distributions.
The finite
properties of
the resulting
estimators are
investigated and
compared with
the
nonparametric
kernel
estimators and
with some
existing
estimators via
simulation
techniques. The
results show
that the
performance of
the proposed
estimators is
better than the
kernel
estimators for
all considered
cases.
Keywords: Matusita
Overlapping
Coefficient;
Maximum
Likelihood
Method;
Parametric
Method; Normal
Distribution;
Relative Bias
and Relative
Mean Square
Error.
|
