On Estimation
of Entropy Value for an Organization
by
S.
E. Omosigho
Department of Mathematics, University
of Benin, Benin City, Nigeria.
and
A. Q. Osagiede
Institute of Education, University of
Benin, Benin City, Nigeria
Abstract
The paper considers the use of entropy
to measure the stability of workforce
of an organization by making use of mainly
wastage data of Edo State Teaching Service
to estimate the wastage probability of
each length of service interval. It is
found necessary to stuffy the degree of
experience present in Edo State Teaching
Service due to the present economic situation
and the state government embargo on employment
in recent years. The study also modifies
the log-normal model of Chu and Lin (1994)
to calculate the wastage probabilities
to tenure classes. Entropy is also calculated
based on this modified log -normal wastage
rates (probabilities) and a comparison
is made between the two entropy values.
Bowey’s stability curves are drawn
for the actual wastage proportions (probabilities)
of PPEB and the modified log-normal wastage
proportions to authenticate the entropy
results.
back to top
Inventory Replenishment
Policy when Demand is Quadratic: Direct
and Inverse Methods
by
F.
E. U. Osagiede and S. E. Omosigho
Department of Mathematics, University
of Benin, Benin City Nigeria
Abstract
We present two new methods for solving
the inventory replenishment problem when
demand is given by over
a finite time horizon H. No shortage is
allowed in the system. The replenishment
times where
are obtained and the order quantity at
each replenishment time is derived. Examples
are given to illustrate these methods.
Keyword: inventory model,
replenishment times, quadratic demand.
back to top
Time Series Analysis
of Malaria Data
by
N.
Ekhosuehi and S. E. Omosigho
Department of Mathematics, University
of Benin, Benin City, Nigeria
Abstract
In this paper, we consider the time series
data on reported cases of malaria in a
large hospital. The data is analysed and
a time series model for the data is proposed.
The models provide a forecasting framework
for planning purpose. The proposed model
is a variant of the Box-Jenkins (ARMA
(p, q)) models. The model identification
is carried out using the correlogram and
the partial correlogram of the time series
data.
Keywords: correlogram,
partial correlogram, model identification
and diagnositc checking. Time series
back to top
An Application
of Homogeneous and Non-Homogeneous Markov
Chains
by
A.
A. Osagiede and S. E. Omosigho
Department of Mathematics, University
of Benin, Benin city, Nigeria
Abstract
The present in application homogeneous
and non-homogeneous Markov chain models
to projecting school enrolment structures.
These models are similar to Markov chain
models used in manpower planning .the
methods used in estimating the parameters
of manpower planning models are adapted
to the data available on the flow of pupils
into, through and out of the system. The
work is illustrated with a case study.
back to top
Work Force Stability
and Suitability
by
S.
E. Omosigho and A. A.. Osagiede
Department of Mathematics, University
of Benin, Benin City, Nigeria
Abstract
The paper examines workforce stability
and suitability. The current measure of
the stability of workforce is entropy.
There are various versions of this measure.
The various versions of entropy assume
that the size of the workforce is constant.
However, this assumption is difficult
to attain in practice. Besides, entropy
does not consider the desired size of
workforce for an organization. Hence,
the present paper proposes augmented ratio
of squares and the recalibrated augmented
ratio of squares to determine the stability
and desirability of an organizational
workforce. These two new measures have
major advantages over the entropy measure.
Keywords: workforce stability,
augmented ratio of squares, entropy.
back to top
A simple algoraithm
for the inventory problem with a linear
incasing demand
by
F.
E. U. Osagiede and S. E. Omosigho
Department of Mathematics, Univeristy
of Benin, Benin City, Nigeria.
Abstract
We consider the inventory system where
a product has linear increasing demand
over a finite horizon and no shortages
are allowed. A new direct search method
is proposed. The search method is implemented
by a compuer program written in MATLAB
and implemented on a PC. Numerous test
problems taken from the literature have
baeen solved using the computer programme.
The results are accurate and the computer
processing time for all the test problems
was found to bse less than one second.
Unlike previous method, the proposed
method is easy to implement.
back to
top
Effect Of Box-Cox
Transformation Parameter On Forecast
From ARIMA(p,1,0) models.
by
N.
Ekhosuehi and S. E. Omosigho
Department of Mathematics, University
of Benin, Benin City, Nigeria.
Abstract
In this paper we consider the effect
of applying a family of Box-Cox transformations
to time series data that exhibit stochastic
upward linear trend which admits differencing
of order one. We examine the interaction
between Box-Cox transformation and forecasting
from the ARIMA(p, 1, 0) model. We demonstrate
the importance of the Box-Cox transformation
parameters =1,
=
0.5 and =
0.
Key words:
Box-Cox transformations; Model fitting;
Forecasting.
back to
top