Modelling Vol. 13, No. 10, pp. 85-100, Press plc. Printed in Great Britain
1990
Math1 Comput. Pergamon
MATHEMATICAL
AND
COMPUTER
MODELLING
REPORTS
This section aims to provide our readers with the title, abstract and other essential data concerning mathematical models that were reported on in other journals, and as selected by our editors. Our readers are encouraged and requested to submit for publication reports of additional mathematical models, in the form here presented. Our own survey extends only to scientific journals published by Pergamon Press; although our readers are invited to contribute from any other source by submitting a tearsheet or good-quality camera-ready copy of the title page including the abstract direct from the journal concerned, providing permission to reproduce has been obtained. Please send all such submissions to Dr D. N. P. Murthy, Department of Mechanical Engineering, University of Queensland, St Lucia, Qld 4067, Australia.
Journal
of Statistical
Planning
EXPERT IMPERFECT
and Inference Vol. 20, pp. 245-277,
SYSTEM MODELS KNOWLEDGE:
1988
FOR INFERENCE A COMPARATIVE
WITH STUDY
AMBROSE GOICOECHEA School for Information
Technology
and Engineering,
Fairfax, VA 22030, Abstract-This on imperfect Cohen’s
paper presents a detailed comparative knowledge:
system
presented
probabilities,
leading methods
Purposely,
the assumptions,
and (6) a class of nonmonotonic
reasoning methods.
Each method
(5) is
and a list of strengths and
the same numerical example is addressed by each method such that we are able to highlight
and computational
Words-Inference
for reasoning based
theory, (3) fuzzy set theory, (4) MYCIN Model,
and discussed in terms of theoretical content, a detailed numerical example,
limitations.
Key
study of six major,
(1) Bayes’ rule, (2) Dempster-Shafer
of inductive
George Mason University,
U.S.A.
requirements
models, expert systems,
that are specific to each method
imperfect knowledge, uncertainty,
in a consistent
manner.
decision support systems,
infer-
ence network, evidential reasoning.
Biometrics
Vol. 44, pp. 635-655,
1988
THE PREVALENCE OF MALARIA IN GARKI, NIGERIA: DOUBLE SAMPLING WITH A FALLIBLE EXPERT JERRY NEDELMAN Department
of Mathematical
Clemson, Abstract-Data of Plasmodium
Sciences, Clemson
South Carolina
2963.4-1907,
University,
U.S.A.
from the Garki Project are analyzed to assess how misdiagnosis falciparum.
Three double-sampling
affects the estimated
prevalence
models that account for the fallibility of the expert are derived Typeset 85
by A,#-‘QX
MATHEMATICAL AND
86
COMPUTER MODELLING REPORTS
and applied. The models incorporate information about the density of parasites in the blood to varying degrees. The error in the estimation of prevalence is quantified; and its dependence on calendar time, age, prevalence, and density is investigated. Prevalence and average density are discovered to be good predictors of the error, with the latter being better. Implications of the double-sampling models for the design of epidemiological surveys similar to the one in Garki are investigated. Key Words-Misclassification,
Stand.
Actuarial
misdiagnosis, plaamodium falciparum.
.I. pp. 129-137, 1988
A RATING
MODEL
FOR
LOSS
OF PROFITS
INSURANCE
TERJE AVEN Rogaland University Center and Hanne Myre, University of Oslo Abstract-In this paper we present a rating model for loss of profits insurance for a production system consisting of n production units. Explicit expressions for the company’s long run expected average claims expenditures are derived. A numerical example is given. Key Words-Loss of profits insurance, indemnity period, long run expected average claims expenditures, SemiMarkov model, regenerative process.
Biometrics
Vol. 44, pp. 787-801, 1988
A RANDOM
EFFECTS MODEL FROM DEPENDENT
FOR BINARY SAMPLES
DATA
MARY COFFEY Department of Mathematical
Sciences, Oakland University, Rochester, Michigan 48063, U.S.A.
Abstract-A model is proposed for the binary data analog of the paired t situation. In this model the success probabilities vary across strata and have a specified probability distribution. The parameters can be estimated by moments or likelihood methods; various large-sample tests are proposed. The esimates and tests do not require fixed sample sizes. The model can be applied to case-control studies in which at least some cases have multiple controls and to other nested designs that have two treatments per stratum and small samples in each treatmentstratum combination. Key Words-Case-control
Znt. J.
Systems
Sci.
studies, categorical data, extra-binomial
variation.
Vol. 19, No. 11, pp. 2291-2306, 1988
DIAGNOSIS
MODEL FOR DYNAMICAL PART 1. DISCRETE SYSTEM
SYSTEMS
YOSHITERU ISHIDA and HIDEKATSU TOKUMARU Division of Applied Systems Science, Faculty of Engineering, Kyoto University, Kyoto 606, Japan Abstract-Many static diagnosis models, which express the static relation between faults and syndrome, have been studied extensively. However, it is difficult to apply these static models to dynamical systems in which the syndrome changes dynamically. In order to construct diagnostic systems for dynamical systems, a dynamic
Mathematical
diagnosis
model
that
propose
a failure
system.
Two
designed
can express
propagation
concepts
Biometrics
Vol.
the dynamic
model
of dynamic
to evaluate
which
705-715,
MODEL
relation
explicitly
diagnosabilities
these diagnosabilities
44, pp.
and Computer
Modelling
of fault propagation expresses
are defined
of the model
87
Reports
should be studied.
some dynamic for the model.
as well as to locate
aspects
In this paper,
of failure
Algorithms a primary
propagation
are presented failure
from
we in a
which are
a syndrome.
1988
SELECTION
AND VALIDATION WITH INTERACTION
FOR YIELD
TRIALS
HUGH G. GAUCH, JR. Department Abstract-The
additive
analysis
of variance
analysis
(PCA)
of Agronomy,
main
(ANOVA)
model
given
a large
trial
to two-way
to the residual
University,
data,
from the additive
of the more familiar
genotype-by-environment
and postdictive
Cornell
and multiplicative
model
trial data is a useful extension predictive
effects
model, Model
14853, U.S.A. model
first
the multiplicative
that is, to the interaction. PCA,
and linear
selection
using data splitting
and
and F-tests,
validation
the additive components
AMMI
regression
respectively.
applies prinipal
analysis
procedures,
are considered A New York
of yield
particularly from
both
soybean
yield
serves as an example.
Z
Analysis
of variance,
components
soybean,
validation.
analysis,
J. Math. Biol Vol.
26, pp. 635-649,
MODELS
Universitlit
P.
biplot,
interaction,
FORMATION
HADELER,
Abstract-Birth, conditions
death,
Tiibingen,
Lehrstuhl
for existence
ZCey Words-Pair
Research
pair formation,
differential
formation,
Policy Vol.
equations.
and global
17, pp. 315-328,
THE
Federal
qualitative
Abstract-This universities,
Change paper mainly
shows that
Research
principal
POPULATIONS
Auf der Morgenstelle
10,
of Germany by a system
of the system
of three nonlinear
are investigated,
homoge-
in particular
the
state. evolution
equation.
DETERMINATION OF RESEARCH IN BRITISH UNIVERSITIES
Centre,
Heriot-Watt
support
and GEOFFREY University,
for restructuring
of research resources.
for such concentration
OUTPUT
WYATT
31-35 Grassmarket,
some of the recent proposals
a process of concentration
there is little
prediction,
1988
begins by noting
through
postdiction,
W&Z-BUSEKROS
are described
homogeneous
PAUL HARE Technological
Republic
properties
of the bisexual
population,
and A.
fur Biomathematik,
and separation
The
stability
two-sex
MODELLING
sequence,
IN BISEXUAL
Ft. WALDST~TTER,
D-7400 Tubingen,
neous ordinary
model
1988
FOR PAIR K.
which
New York (AMMI)
and then applies
ANOVA,
interaction.
perspectives,
Ithaca,
interaction
except
Edinburgh
EHl
research
activity
We then review
in the most
2HT,
the available
resource-intensive
U.K.
in British evidence, disciplines.
MATHEMATICAL AND COMPUTER MODELLINC REPORTS
88
Since there is little &m information about the shape of “research production functions”, either for institutions or for the system as a whole, the core of the paper studies the factors likely to be important in such functions and investigates the consequences of particular functional forms. Finally, it is concluded that more research is needed to gain a better understanding of the research process in universities, and that in the meantime it would be unwise to embark on a wholesale restructuring of the system.
Int. J. Systems
Sci. Vol. 19, No. 11, pp. 2367-2379, 1988
MODELLING
AND
IDENTIFICATION WITH GENERAL
OF OFFSHORE DAMPING
TOWERS
K. A. MOUSTAFA Department
of Design and Production
Engineering, Faculty of Engineering, Aim-Shams University, Abbassia, Cairo, Egypt K. ASFAR and F. DRIES
Department of Mechanical Engineering, Jordan University of Science and Technology, Irbid, Jordan Abstract-The modelling problem of offshore towers is considered. A model is proposed which includes viscous and hydrodynamic drag damping and takes into account the stochastic nature of the problem. The model is decoupled into a set of modal equations by using the complex mode theory. A technique is developed to identify the modal parameters and the complex mode shapes by using the noisy excitation and response measurements. This technique is computationally efficient and accurate in the sense of minimizing a criterion function of the estimates variance. A simulation example is presented to demonstrate the results.
J. Appl.
Prob.
Vol. 25, pp. 492-500, 1988
AN IMPROVED
STOCHASTIC AMONG
MODEL ALLOWING INDIVIDUALS
INTERACTION
P. B. CERRITO Department
of Mathematics,
University of South Florida, Tampa, FL 33620, USA
Abstract-The purpose of this paper is to explore the properties of a new model which can describe the interaction among individuals. Various other models have been proposed, but all assume that each individual acts independently of others. Our model permits the assumption that individuals can be influenced and dominated by other individuals. We also demonstrate how this model can describe the interaction of various species in an ecosystem. This model will lead naturally to the definition of a population index of diversity. A means of estimating the values of the model will also be given. ICey Words-Markov
chains, behavior interaction, diversity index, estimation, regular semigroup
Mathematical
Int.
J. Systems
SC;. Vol.
19, No.
and Computer
11, pp. 2307-2319,
Modelling
Reports
89
1988
DIAGNOSIS MODEL FOR DYNAMICAL SYSTEMS PART 2. CONTINUOUS SYSTEM ISHIDA and HIDEKATSU TOKUMARU
YO~HITERU Division of Applied Abstract-While
Systems
most existing systems
are static.
Consequently,
dynamical
systems.
Although
model of differential equations
this model-based systems.
in qualitative
discussed.
theory.
The distinguishability
are derived by these qualitative
Journal
of Dynamics,
BOND
We interpret these qualitative
Strategies for monitoring
Abstract-The
physical
systems
products
generalizing
We propose a diagnosis which has been
stabilities of the model
is also studied with a new concept
of invariant
as well as for the diagnosis of the dynamical
Engineering,
modeling
University
IN FINITE
systems
variables and transformer
the power flow and modulation
of Texas, Austin,
deformation.
Texas 78712 to represent the hysteretic
Second and fourth order tensors rep
or gyrator moduli, respectively, calculations
STRAIN
WV
theory of bond graphs may be employed
solids in finite strain elastic-plastic
resent stress power conjugate
mode.
and Control Vol. 110, 1988
of Mechanical
of continuous
diagnosis of
results in terms of the diagnosis
MODELING OF CONTINUOUS SOLIDS ELASTIC-PLASTIC DEFORMATION
Department
for diagnosis
results.
Measurement
GRAPH
proposed
it is not used much in practice
The dynamic aspects of the qualitative
of the sign pattern
E. P. FAHRENTHOLD and A.
deformation
most models
606, Japan
The model can be expressed by a signed directed-graph,
matrix
fault pattern defined in the model.
Kyoto
is used for a model-based
diagnosis has been well studied,
and also present new results for the model.
are mainly
University,
difficulty and the difficulty of obtaining the mathematical
model for continuous dynamical studied extensively
Kyoto
requiring diagnosis are dynamical,
a mathematical
owing to the computation
model,
Science, Faculty of Engineering,
with corresponding
of scalar and vector bond graphs.
inner
Capacitors
and resistors of a tensor type suitably represent strain energy storage, associated flow rules in plasticity, and other familiar concepts.
An important
application
of this simulation
technique arises in constitutive
modeling
studies
of nonlinear materials.
Journal
of Dynamic
Measurement,
Syslems,
MODELING
AND
and Control
SIMULATION
RAYMOND MERALA, Mechanical Abstract-A
dynamic
Engineering
configuration.
MONT
Department,
model is developed
tively arbitrary geometric
Vol.
110, 1988
OF A SUPERCHARGER
HUBBARD,
and TAKASHI MIYANO
University of California,
for simulating
and predicting
A thermodynamic
Davis, CA 95616
performance
used to derive continuity and energy equations linking the various control volumes. and understand
the casual implications
Heat transfer is neglected.
Simulation
for superchargers
of laws governing flows between
Bond graphs also serve to study
control volumes
outputs include time histories of pressure,
with each control volume, time histories of the various flows in the supercharger,
efficiency.
Volumetric
within three percent of experimentally supercharger
performance
measured
over a wide range of speed/pressure values.
The simulation
to several key design parameters,
and side plate clearance distances.
including
and system
temperature,
associated
efficiencies are predicted
of rels
control volume approach and bond graph models are dynamics. and energy
and overall volumetric
ratio combinations
is used to investigate
rotor-rotor
mass,
separation,
and are
the sensitivity
of
and rotor-housing
MATHEMATICAL AND COMPUTER MODELLINC REPORTS
90
Journal of Dynamic
MODELING
Systems,
AND
Measurement,
and Control Vol. 110, 1988
IDENTIFICATION SYSTEMS WITH
OF A CLASS OF SERVOMECHANISM STICK-SLIP FRICTION
KA C. CHEOK, HONGXING Hu, and NAN K. LOH Center for Robotics
and Advanced Automation,
Oaklaud University, Rochester, MI 48309
Abstract-This paper describes a technique for modeling and identifying a class of nonlinear servomechanism systems with stick-slip friction. The physics of the stick-slip friction is considered in modeling the process. Identification of the system parameters is formulated as a nonlinear optimization problem. A modified simplex algorithm is proposed as the optimization procedure. The difficulties encountered in choosing identification algorithm and input signals for the problem are discussed. A simulation example of a servomotor system is provided.
Mathematical
Biosciences
Vol. 92, pp. 119-199, 1988
MODELING THE
AND ANALYZING HIV TRANSMISSION: EFFECT OF CONTACT PATTERNS JOHN A. JACQUEZ
Departments of Biostatics and Physiology, The University of Michigan, Ann Arbor, Michigan CARL P. SIMON Departments of Mathematics,
Economics, and Public Policy, The University of Michigan, Ann Arbor, Michigan JAMES KOOPMAN
Department of Epidemiology,
The University of Michigan, Ann Arbor, Michigan LISA SATTENSPIEL
Department of Anthropology,
University of Missouri, Columbia, Missouri
TIMOTHY PERRY Departments
of Physiology
and Internal Medicine, The University of Michigan, Ann Arbor, Michigan
Abstract-A compartmental model is presented for the spread of HIV in a homosexual population divided into subgroups by degree of sexual activity. The model includes constant recruitment rates for the susceptibles in the subgroups. It incorporates the long infectious period of HIV-infected individuals and allows one to vary infectiousness over the infectious period. A new pattern of mixing, termed preferred mixing, is defined, in which a fraction of a group’s contacts can be reserved for within-group contacts, the remainder being subject to proportional mixing. The fraction reserved may differ among groups. In addition, the classic definition of reproductive number is generalized to show that for heterogeneous population in general the endemic threshold is PDcy, where cy is the mean number of contacts per infective. The most important finding is that the pattern of contacts between the different groups has a major effect on the spread of HIV, an effect inadequately recognized or studied heretofore.
Mathematical
Biosciences
Vol. 92, pp. 201-229, 1988
A METHODOLOGICAL STUDY OF AN AIDS
OF A STOCHASTIC EPIDEMIC
MODEL
CHARLES J. MODE, HERMAN E. GOLLWITZER, and NRA HERRMANN Department of Mathematics
and Computer Science, Drexel University, Philadelphia,
Pennsylvania 19104
Abstract-A model of an AIDS epidemic in a population of male homosexuals was formulated as a stochastic population process. The paper is a methodological study in the sense that computer-intensive methods were used to investigate some properties of the model statistically rather than relying solely on classical methods of deductive mathematics. Three factors of importance in the evolution of an AIDS epidemic were studied in a numerical factorial experiment. These factors were the distribution of the latent period of HIV, the probability of
Mathematical
and Computer Modelling Reports
91
infection with HIV per sexual contact with an infected individual, and the distribution of the number of contacts per sexual partner per month. The numerical experiment suggested that the distribution of the latent period of HIV will have a decisive impact on the evolution of an AIDS epidemic but this impact will depend crucially on the levels of the other two factors. A Monte Carlo experiment suggested that if forecasts of an epidemic were made solely on the basis of deterministic nonlinear difference equations embedded in the stochastic population process, then prediction of the number of individuals infected with HIV and AIDS cases may be overly pessimistic.
SIAM
J. Appl.
Math.
Vol. 48, No. 6, 1988
MARTINGALE
APPROACH
FOR MODELING
DNA
SYNTHESIS
A. GERARDI and G. NAPPO Dipartimento
di Matematica,
Universitb di Roma “La Sapiensa,” P. Pe Aldo More, 00100 Rome, Italy
Abstract-The relation between continuous and discrete models for DNA synthesis is treated. A law of large numbers is given when cells can die during the mitotic period. In the critical case the fluctuations around an equilibrium state are studied. In particular the convergence to an Ornstein-Uhlenbecktype process is proved. Key
Word.+-Martingales,
J. Math.
weak convergence, nuclear spaces.
Biol. Vol. 27, pp. 29-48, 1989
STATISTICAL
MODELLING
OF MITOCHONDRIAL
POWER
SUPPLY
A. T. JAMES and J. T. WISKICH Departments of Statistics and Botany, The University of Adelaide, Adelaide, S.A. 5000 Ft. A. J. CONYERS Department of Biochemistry, Alfred Hospital, Melbourne, Victoria 3181, Australia Abstract-By experiment and theory, formulae are derived to calculate the response of mitochondrial power supply, in flux and potential, to an ATP consuming enzyme load, incorporating effects of varying amounts of (i) enzyme, (ii) total circulating adenylate, and (iii) inhibition of the ATP/ADP translocclse. The formulae, which apply between about 20% and 80% of maximum respiration, are the same as for the current and voltage of an electrical circuit in which a battery with potential, linear in the logarithm of the total adenylate, charges another battery whose opposing potential is also linear in the same logarithm, through three resistances. These resistances produce loss of potential due to d&equilibrium of (i) intramitochondrial oxidative phosphorylation, (ii) the ATP/ADP translocase, and (iii) the ATP-consuming enzyme load. The model is represented geometrically by the following configuration: when potential is plotted against flux, the points lie on two pencils of lines each concurrent at zero respiration, the two pencils describing the respective characteristics of the mitochondrion and enzyme. Control coefficients and elasticities are calculated from the formulae. Key Wo&-Mitochondria,
respiration, ATP/ADP
ratio, kinetics, thermodynamics, ,data-modelling
MATHEMATICAL AND
92
J. Math.
Biol.
COMPUTER MODELLING REPORTS
Vol 27., pp. 49-64, 1989
AN EPIDEMIOLOGICAL NONLINEAR
MODEL WITH A DELAY INCIDENCE RATE
AND
A
H. W. HETHCOTE Department of Mathematics,
University of Iowa, Iowa City, IA 52242, USA M. A. LEWIS
Centre for Mathematical
Biology, Mathematical
Department of Mathematics,
Institute, 24-29 St. Giles, Oxford OX1 3LB, UK and
University of Victoria, B.C. V8W 2Y2, Canada
P. VAN DEN DRIESSCHE Department of Mathematics,
University of Victoria,
B.C. V8W 2Y2, Canada
Abstract-An epidemiological model with both a time delay in the removed class and a nonlinear incidence rate is analysed to determine the equilibria and their stability. This model is for diseases where individuals are first susceptible, then infected, then removed with temporary immunity and then susceptible again when they lose their immunity. There are multiple equilibria for some parameter values, and, for certain of these, periodic solutions arise by Hopf bifurcation from the large nontrivial equilibrium state. Key Words-Epidemiological
J. Math.
Biol.
model, Hopf bifurcation, nonlinear incidence, time delay.
Vol 27., pp. 81-90, 1989
NUMERICAL
SIMULATION OF A STOCHASTIC MODEL CELLS SUBMITTED TO CHEMOTHERAPY
FOR
CANCEROUS
M. ABUNDO and C. Ross1 Dipartimento
di Matematica,
11Universita di Roma, Via Orazio Raimondo, I-00173 La Romanina (Roma),
Italy
Abstract-A stochastic model is proposed to study the problem of inherent resistance by cell populations when chemotherapeutic agents are used to control tumor growth. Stochastic differential equations are introduced and numerically integrated to simulate expected response to the chemotherapeutic strategies as a function of different parameters. Satisfactory demonstration runs of the model indicate that it could represent a useful tool in verifying the results of experimental and clinical chemotherapy courses and planning treatment strategies. Some types of behaviour are illustrated graphically. ZCey Word.+-Cell
J. Math.
Biol.
LOCAL
resistance, stochastic models, simulation.
Vol 27., pp. 65-80, 1989
VS. NON-LOCAL
INTERACTIONS
IN POPULATION
DYNAMICS
J. FURTER and M. GRINFELD Department of Mathematics,
Heriot-Watt
University, Riccarton, Edinburgh, EH14 4AS, Scotland, UK
Abstract-In this work we examine two models of single-species dynamics which incorporate non-local effects. The emphasis is on the ability of these models to generate stable patterns. Global behavior of the bifurcating branches is also investigated. Key Words-Reaction-diffusion
equations, non-local effects, stable patterns.
Mathematical J. Math.
and Computer Modelling Reports
93
Vol 27., pp. 105-113, 1989
Biol.
ON STOCHASTIC
COMPARTMENTAL
MODELS
Rudjer Bor(kovif. Institute, POB 1016, Yu-41001 Zagreb, Yugoslavia Abstract-The statistical averaging of compartmental models with parameters being random processes is derived for the case of vanishing input and uniformly bounded input. the difference of resulting models is discussed. Key Words-Compartmental
Appl.
Math.
Modelling
models, stochastic models.
Vol 13., 1989
MATHEMATICAL
MODELLING OF STIMULUS-SECRETION COUPLING IN THE PANCREATIC B-CELL VI. CELLULAR HETEROGENEITY AND RECRUITMENT WILLY J. MALAESE and ANN OWEN
Laboratory
of Experimental
Medicine and Department of Endocrinology,
(Erasmus Hospital),
Brussels Free University, Brussels, Belgium Abstract-Insulin release from the pancreatic B-cell may display rhythmic oscillations between basal and stimulated values, reflecting comparable fluctuations in the rate of Ca *+ inflow. The mathematical modelling of such a process allows simulation fo the biphasic pattern of insulin release and its dependence on glucose concentration, when allowance is made for the heterogeneity of both the time course of secretory events and threshold glucose concentration for stimulation of insulin release in distinct B-cells. It is proposed that such a heterogeneity and the consequent recruitment of cells at increasing concentrations of the hexose indeed play an essential role in the dynamics of hormonal release. Key
Worda-Insulin
Optimal
Control
secretion, mathematical
Applications
MODELLING
Ed Methods
AND CONTROL
modelling, pancreatic islet.
Vol lo., pp. 2138,
1989
OF A ROTARY
CRANE WITH A FLEXIBLE
JOINT
TADASHI YOSHIMOTO and YOSHIYLJKISAKAWA Department of Control Engineering, Faculty of Engineering Science, Osaka University, Toyonaka, Osaka 560, Japan Abstract-A dynamical model is derived for the control of a rotary crane, which has a flexible joint and makes three kinds of motion (rotation, load hoisting and boom hoisting) simultaneously. The goal is to transfer a load to a desired place in such a way that at the end of the transfer the swing of the load decays as quickly as possible. A new open-loop plus feedback control scheme is proposed.
Key Words-Flexible
rotary crane, open-loop plus feedback control, observer, stabilization of swing
MATHEMATICAL
94
Quality
and Reliability
Engineering
A MULTI-FAILURE
RAPONE,
system
considers
mathematical
by applying
WordsTransit
system,
SIMPLE
TWO
and GIANPAOLO Motori,
1 80125 Naples,
operating
a new
transit
may be delayed are compared
to a common
service
dependability,
Models
Vol 5., No.
DIMENSIONAL
literature
as an effective policy
model
with those relative
from
failure which
in a trip.
by using the Monte
to different
of
is proposed,
of more than one failure are overcome
measure
and maintainability
and recovery
dependability
variables
Council,
with reliability
characteristics
by the occurrence
of random
Research
Italy
reliability
system
FOR TRANSIT
PULCINI
National
link system dependability
special
model
the model
Statist.-Stochastic
Common.
REPORTS
1989
in the transportation
with the algebra
The results of the proposed
Istituto
e Matteucci,
In this paper
associated
CALABRIA
models
incorporating
system.
that a passenger
difficulties
the literature, Key
transit
the possibility
method.
Dependability
of subsystems,
of each particular
Barsanti
has been recognized
service quality.
characteristics
RAFFAELA
and Reliability,
Piazza
transit
Vol 5., pp. 47-52,
International
of Statistics
Abstract-Dependability
MODELLING
ADDITIVE DEPENDABILITY MODEL SYSTEM EFFECTIVENESS ANALYSIS
MARIO Department
COMPUTER
AND
modelling
The Carlo
approaches
in
test scenario. Monte
1, pp.
Carlo
131-159,
MODELS
simulation.
1989
FOR THE
SPREAD
OF A DISEASE
DAVID GREENHALGH Control
and Management
Systems
Division,
Cambridge
Cambridge, Abstract-A a fixed
simple
population.
assumptions
versions
counterexamples
Prob.
Vol.
the
the optimal
25., pp. 649-662,
Abstract-A
drug addiction Expressions
also developed. a competing Key
This
death
of Statistics,
Mill Lane,
to model
which
Chain Binomial
make
epidemic
of people
the spread assumptions models.
infected.
of a disease
The epidemic
Some
amongst
corresponding
results,
to the
is controlled
conjectures
process
The
ADDICTION and P. W. University
in which individuals
drug addiction
A.
MODEL DAYANANDA
of Georgia,
Athens,
GA
in a closed population
and factorial
moments
process can be viewed
30602, USA can become
are obtained.
as a particular
addicts
A generalized
or pushers model
addicts,
pushers,
state probabilities,
factorial
moments,
is
case of what may be called
process.
Worda-Susceptibles,
and
are presented.
for the state probabilities
generalized
Department,
England
is proposed
number
Engineering
1988
L. BILLARD
is modelled.
lRX,
are obtained
expected policy
A DRUG
Department
model
and Reed-Frost
so as to minimize concerning
stochastic
of this model
made in the Greenwood
by immunization
J. Appl.
two dimensional Two
CB2
University
competing
death
process.
Mathematical
Reliability
Engineering
and
AN ACCELERATED
Vol 24., pp.
Safety
System
LIFE
and Computer
TEST
Modelling
223-230,
MODEL
Department
of Industrial
Engineering,
Korea
Advanced
Chongryang, Abstract-Constant Weibull
and progressive
lifetime
distributions
prior distribution to estimate
and other parameters
the model
of Marketing
Jou~sal
in which
Institute
life
constants.
POWER
LAW
and Technology,
PO
Box 150,
of Korea based
on the
inverse
power
law
and
scale parameter at use condition has a gamma Bayes and maximum likelihood methods are used
parameters.
Monte
Carlo study is performed
Vol 53., pp.
80-88,
1989
MODELING
INVERSE
of Science
test models
the Weibull
are unknown
THE
CHUNG
Seoul, Republic
stress accelerated
are proposed
95
1989
WITH
D. S. BAI and S. W.
Reports
to investigate
THE DECISION TO ADD NEW BY CHANNEL INTERMEDIARIES
the behaviour
of the estimators.
PRODUCTS
VITHALO R. RAO Johnson
Graduate
School of Management,
Cornell
University
EDWARD W. MCLAUGHLIN Department Abstract-Using logistic
data collected
regression
how different validation
variables system
on new products
to describe
influence
performance.
the marketing
Journal
models
of Agricultural
presented
the intermediary’s
those decisions.
Implications
Economics,
The
Cornell
to a major
accept/reject logistic
model
of these results for marketing
University
channel intermediary,
the authors
decisions for those products. is shown
to fit extremely
strategies
estimate
Results indicate
well with excellent
and for improving
performance
of
are discussed.
of Forecasting
Vol 8., pp.
A DYNAMIC
85-96,
LINEAR
1989
MODEL TIME
APPROACH FOR DISAGGREGATING SERIES DATA
M. AL-OSH King Abstract-An aggregate
approach series, it,
is proposed
of k period
problem
in a dynamic
filtering
techniques.
linear
An
from
the observed
model
ICey
Word+--Aggregation,
hood
estimation.
Saud University,
for obtaining
non-overlapping
model
form.
ad hoc procedure of the aggregates. d&aggregation,
Then
estimates
of the basic (disaggregated)
sums of Zi’s is available. estimates
is introduced
process,
The
approach
of Zi can be obtained
for deriving
An application ARIMA
Saudi Arabia
a model
of this approach dynamic
linear
series, Zi, when only an is based on casting
by application
the
of the Kalman
form for the unobserved
basic series
to a set of real data is given. model,
state
space,
maximum
likeli-
MATHEMATICAL AND
96
Journal
of Forecasting
TIME
Vol 8., pp.
SERIES
75-83,
COMPUTER MODELLINC REPORTS
1989
FORECASTS
AND
EXTRA-MODEL
INFORMATION
ALAN PANKRATZ Department Abstract-Often
of Economics
and Management,
a forecaster has supplementary
that cannot be included
DePauw University,
information
directly in a time series model.
Greencastle,
(e.g. field reports or forecasts from another source)
Especially
interesting
are cases where this information
is given at time intervals that are different from those of the time series model forecasts. considered a numerical and a model-based with ARIMA regression
model forecasts.
statistical
Key
(transfer function)
Words-ARIMA
European
Journal
for combining
This paper extends both methods model forecasts.
arises as a special case of the model-based
transfer function,
method
extra-model
to vector ARMA
Previous
authors have
information
of this type
model forecasts and dynamic
It is also shown that a Lagrange multiplier
procedure.
models, benchmarking,
IN, USA
numerical
procedure
dynamic
regression,
An empirical example is given.
composite
forecasts,
distributed
lag regression,
vector ARMA.
of Operational
Research
Vol 39., pp. 231-242,
COMPETITIVE
SPATIAL H. A.
Faculty of Administration,
University
1989
MODELS
EI~ELT
of New Brunswick,
Fredericton,
New Brunswick,
Canada
E3B 5A3
G. LAPORTE Centre de recherche sur les transports, Mont&al, Abstract-A models
de MontrBal, C.P. 6128,
actions.
Succursale
A,
QuCbec, Canada H3C 357
large number of decision situations can be modeled as spatial problems.
allow only one firm or facility to locate,
competitors’
Universitt
competitive
models permit
Whereas
the other facilities
the usual location to react to their
One question is then not only where an individual facility will locate but if repeated relocation
of the facilities involved does converge towards a predictable described in the literature.
pattern.
Many competitive
spatial models have been
In this review we first formalize the general model and delineate its components.
we survey some of the major results and put them into perspective.
We conclude
Then
the review with a number
of
research ideas which may be explored in the future. Key
Words-Competitive
models,
location
problems,
Nash
equilibria,
Hotelling
models,
spatial
optimization,
social optima.
Appl.
Statist.
Vol 38., No.
A PROBABILISTIC
1, pp. 95-110,
MODEL
1989
OF SQUASH:
STRATEGIES
AND
APPLICATIONS
JOHN SIMMONS Department Abstract-Optimal
of Physics and Astronomy,
strategies
chain. Further implications
Glasgow University,
for squash are presented
assuming
Glasgow,
G12 SQQ,
that the game can be modelled
of this model are discussed in relation to handicapping
The validity of the model is briefly discussed in the light of data.
Key
Markov
chains, scoring rules, squash.
a~ a Markov
and the correspondence
different scoring systems. WordtEstimation,
UK
between
Mathematical
Appl.
Statist.
Vol 38., No.
1, pp.
l-50,
and Computer
Modelling
Reports
97
1989
SPACE-TIME MODELLING WITH LONG-MEMORY DEPENDENCE: ASSESSING IRELAND’S WIND POWER RESOURCE JOHN HASLETT Department
of Statistics,
Trinity
ADRIAN University Abstract-We
consider
a site for which
few data
meteorological which
short-memory kriging,
of Washington,
Seattle,
USA
on wind
the main
features
of wind
autocorrelation
modelling
estimator,
2, Ireland
of the long term average power output from a wind turbine generator at Long term records of wind speeds at the 12 synoptic speeds are available. Inference is based on a simple and parsimonious approximating model are also used.
temporal
ARMA
as a point
for
Dublin
estimation
stations
accounts
College,
E. RAFTERY
speeds
in Ireland,
and long-memory
and fractional
and good interval
differencing estimators
temporal
in a natural
namely seasonal effects, spatial correlation, It synthesizes deseasonalization, dependence. way. A simple kriging estimator performs well
result from the model.
The resulting
procedure
is easy to apply
in practice. I
W&P-Deseasonalization,
Appl.
Statist.
fractional
Vol 38., No.
1, pp.
differencing,
127-138,
MULTIPLE-SPELL
kriging,
REGRESSION
of Statistics,
Abstract-General
models
how the successive special
cases of the general
timescales
MODELS
without
model
common
further
analysing
programming.
Icey Words-Completely
Finally,
and partially duration
Appl.
Vol 13., 1989
Modelling
TWO-EQUATION
Abstract-The
derivation
of conservation and
the
corresponding
constraints closure that
imposed
assumptions)
the model
are considered.
Konstanz, A general
stochastic
of different
DATA
such as Markov
models
are applied
models,
different
hazard
theory
Germany which
indicates Various
are investigated:
processes
different
or semi-Markov
GLIM
and RATE
to the duration
timescales,
West
process is presented.
timescales
data such as SAS, BMDP,
transition-specific
of a two-equation balance
balance
processes. can be used
of unemployment,
and
and conclusions.
maximum
likelihood
estimation,
rate.
laws
for
to within
model
conservation
the mean
turbulent
stress tensor,
of predicting
VALENTINE
Clarkson
of energy, buoyant
NY,
is presented.
and entropy
flow
field
are used to formulate heat
Potsdam,
USA
with the second law of thermodynamics
approximation
the heat flux vector,
nongradient
University,
consistent
the Boussinesq
law of thermodynamics
for the turbulent
and D. T.
Engineering,
turbulence
of momentum,
by the second
has the capability
G. AHMADI,
and Industrial
the effects of buoyancy
of mass,
5560, D-7750
is shown to lead to false interpretations
G. TANNOUS,
of Mechanical
that takes into account
processes
parameterized
model,
DURATION
THERMODYNAMICAL MODEL FOR TURBULENT BUOYANT FLOWS. PART I: THEORY A.
Department
data
The implications
multiple-spell
FOR
by an underlying
for duration
framework
multistate-multiepisode
Math.
duration
stochastic
programs
the data in a single-spell
Postfach
are generated
are discussed.
underlying
computer
persistence.
HAMERLE
of Konstanz,
for multiple-spell
lead to different
Occasionally
University
spells of an individual
interpolation,
1989
ALFRED Department
optimal
inequality
The
are averaged,
are developed.
The
the constitutive
equations
and the energy
diffusion.
The equations
new
flux vector.
model
contains
physical (or
It is shown empirical
MATHEMATICAL
98 constants data.
that are determined
Furthermore,
turbulence Key
with added
SOME
solutions
the model
equations
REPORTS
of limiting reduce
turbulent
flows with experimental
to those of the common
k-c model
for
terms.
buoyant
Modelling
Math.
MODELLING
of analytical
simplifications
buoyancy
Word-Turbulent
Appl.
via comparison
with certain
COMPUTER
AND
flow,
two-equation
model,
fluid mechanics.
Vol 13., 1989
NONLINEAR
OPTIMIZATION MODELLING FOR PLANNING OF LARGE MARKET-ORIENTED SYSTEMS: WITH AN APPLICATION TO REAL HEALTH DATA
OBJECTIVES
R. S. SECALL
Department Abstract-This prediction
paper
of patient
of its mathematical constrained model their
form
Massachusetts of the origin
derivations
hospitals.
with certain constrained derives
on objectives
This
Math.
of Lowell,
article
model
has been
efficiency,
model
States better
and formulations
and accessibility.
This
model
The
to predict predict
the flows
An application
of the equity
optimization
models,
Laboratorio
systems a finite Key
mathematical curve
of differential
equations
associated
bar left initially
Word-Schock
National
model
the stress-strain
of the general
of
assumptions
form.
This paper
programming
objective
from
in the state
models
to real health
equity,
efficiency,
accessibility,
market
systems.
Vol 13., 1989
Modelling
ROGCRIO
which
flows
is presented.
MODELING AND NUMERICAL SIMULATION OF SOME PHENOMENA INVOLVING ELASTIC-NONLINEAR
Abstract-A
constrained
of patients
the underlying
nonlinear
for the bases all
than the destination
than the destination
of several
article
destination
patient
However,
modelling
sequential rather
article.
and Europe
of some variables.
in the United
constrained
in the earlier
USA
mathematical
domain.
used to accurately
to the definition
MA,
presented
form of the gravity
and used in England
fit the situation
the origin
which
Lowell,
a large geographical
constrained
model
modifications
of equity,
Words-Nonlinear
Appl.
within
an origin
validated,
data for the state of Massachusetts Key
of a previous
“80~s”
upon
developed,
to acute
mathematically based
is a continuation
client customer
University
which was used for the basis of the modelling
was initially homes
of Mathematics,
is proposed
Glimm’s
Cientifica,
method,
Rio de Janeiro,
to analyze the dynamical This phenomenon results obtained
Brazil
behavior of elastic-nonlinear bars in is governed by nonlinear hyperbolic
shock waves in their solution.
and some typical
in a nonequilibrium
DA GAMA
points.
that may present problem
SALDANHA
de Computa@o
has no inflection
RiemaM
waves,
MARTINS
DYNAMICAL BARS
We present
numerically
state. nonlinear
elasticity,
elastic rod, vibrations.
the complete
by Glimm’s
solution
method
for
Mathematical and Computer Modelling Reports in the Engineering and Informational
Probability
MODELING
POWER
99
Sciences Vol. 2, pp. 435-459, 1988
OF STOCHASTIC
PETRI NETS FOR SIMULATION
PETER J. HAAS and GERALD S. SHEDLER IBM Ahnaden Research Center, San Jose, California 95120-6099 Abstract-Generalized semi-Markov processes and stochastic Petri nets have been proposed as general franu+ works for a discrete even simulation on a countable state space. The two formal systems differ, however, with respect to the clock setting (event scheduling) mechanism, the state transition mechanism, and the form of the state space. We obtain conditions under which the marking process of a stochastic Petri net ‘mimics” a generalized semi-Markov process in the sense that the two processes (and their underlying general state-space Markov chains) have the same finite dimensional distributions. The results imply that stochastic Petri nets have at least the modeling power of generalized semi-Markov processes for discrete event simulation.
Statistics
in Medicine Vol. 8, pp. 259-262, 1989
AN APPLICATION OF A MATHEMATICAL MODEL FOR TIME LAG IN CASE REPORTING
TO ADJUST
KUNG- JONG Lur Biometrics Branch, Division of Injury Epidemiology and Control, Center for Environmental Health and Injury Control, Centers for Disease Control, Atlanta, GA 36333, U.S.A. RICHARD K. RUDY Gillis W. Long Hansens Disease, Regional Hansens Disease Program, Carville, LA 70721, U.S.A. Abstract-In a dynamic, fluctuating surveillance system, the time lag case of reporting often causes an artificial plateau in an epidemic curve. Arbitrarily ignoring data reported in the most recent period to avoid this bias causes the loss of valuable information. In this report, we propose an application of a mathematical model to adjust for the underreporting bias owing to the time lag of the reporting process. We present an example using the acquired immunodeficiency syndrome incidence data for the homosexual group in the San Francisco surveillance system to illustrate and evaluate prospectively this proposed technique. The results show that the adjusted incidence obtained with the model agrees reasonably well with the true incidence, except for the last month of the period under consideration. ZCey Words-Surveillance,
Appl.
Math.
acquired i-unodeficiency
syndrome, homosexual,
Modelling Vol. 13, 1989
A NUMERICAL
MODEL FOR SIMULATING TWO-PHASE THROUGH POROUS MEDIA
FLOW
ANANT R. KUKRETI and YATENDRA RAJAPAK~A School of Civil Engineering and Environmental Science, The University of Oklahoma, Norman, OK, USA Abstract-A numerical model is developed for flow of immiscible and incompressible fluids through porous media, representing subsurface oil reservoirs. The spatial domain is discretized by using finite element theory, and the element matrices are obtained by Gale&in’s method. The resulting nonlinear equations are solved in the time domain by subdividing the total simulation time into small time increments and approximating the variation of the variables by a polynominal of time within each time increment. The coefficients of the polynominal are determined iteratively. The iterative algorithm is presented for a linear and quadratic approximation. The method is demonstrated for two-phase flow through a homogeneous square domain. Results of a sensitivity study conducted to investigate the effect of the type of element, mesh configuration, size of time increment adopted, and the order
MATHEMATICAL AND COMPUTER MODELLING REPORTS
100
of approximation used on the accuracy, convergence, and computational presented. The results are also compared to those of another method. Key Wordthmniscible,
efficiency of the results predicted
are
flow, two-phase, porous media, finite element.
Statistics in Medicine Vol. 8, pp. 263-266, 1989
A MODEL-BASED
APPROACH TO THE IMPUTATION HOME INJURY INCIDENCES
OF MISSING
DATA:
JUDITH M. CONN, KUNG-JONG LUI and DANIEL L. MCGEE Biometrics Branch, Division of Injury Epidemiology
and Control, Center for Environmental
Health and Injury
Control, Centers for Disease Control, Public Health Service, U.S. Department of Health and Human Services, Atlanta, Georgia, 30333, U.S.A. Abstract-Missing or incomplete data cases are a problem in all types of statistical analyses. In disease surveillance, this problem inhibits determining the actual incidence of a disease event and monitoring the disease occurrence. Several statistical techniques have been developed to impute values for incomplete data cases. We present a model-based approach to the imputation of missing data elements as applied to determining the incidence of home injury deaths. Key
Words-Imputation,
Automatica
incomplete data, logistic regression, partially missing data.
Vol. 25, No. 2, pp. 185-206, 1989
DEVELOPMENT
AND APPLICATION SUPPORT
OF AN INTERACTIVE SYSTEM
MODELING
Y. NAKAMORI Department of Applied Mathematics,
Konan University, 8-9-l
Okamoto, Higashinada-ku, Kobe 658 Japan
Abstract-This paper outlines an interactive modeling support system that helps model building for complex large-scale systems involving hum& behavioral aspects. The system consists of combined modeling techniques using statistical and graph-theoretical approaches, and multi-stage person-computer dialogues. It assists flexible determination of the model structure according to the advice of experts, and clear definition of the model objective with conviction. Three examples are presented to show its applicability to the wide range of concrete problems. I(ey Words-Man-machine interfaces and dialogue systems, mental models, human decision making, decision support systems, computer-assisted modeling.