Heuristic "Optimization": Why, When, and How to Use It

Author(s): Stelios H. Zanakis and James R. Evans

Source: Interfaces, Vol. 11, No. 5 (Oct., 1981), pp. 84-91
Published by: INFORMS
Stable URL: http://www.jstor.org/stable/25060151
Accessed: 21-08-2015 06:02 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.

INFORMS is collaborating with JSTOR to digitize, preserve and extend access to Interfaces.

http://www.jstor.org

This content downloaded from 125.20.9.248 on Fri, 21 Aug 2015 06:02:37 UTC
All use subject to JSTOR Terms and Conditions

INTERFACES
Vol.

11, No.

October

5,

Sciences
Institute of Management
0092-2102/81/1105/0O84S01.25

1981, The

Copyright ?
1981

HEURISTIC "OPTIMIZATION":
WHY, WHEN, AND HOW TO USE IT
Stelios H.
Florida

International

University,

Zanakis
of Business

College

Miami,

Management,

Florida

Administration,

of

Department

33199

and
James R. Evans
University

of Cincinnati,

College

of Business

Administration,
Ohio 45221

Department

of Quantitative

Cincinnati,

Analysis,

for quick, simple, and imple

The pressing need of real-world
problems
on improved
increased
research productivity
coupled with a recently
are rapidly
and
the need, usage,
heuristic methods
evaluated
increasing

Abstract.

solutions

mentable

and rigorously
for heuristics.

tematically
and proper

for heuristic
paper presents a framework
in attitudes
this change
towards heuristics,

This

respect

examining

"optimization"
their desirable

by sys
features,

usage.

"heuristic"
derives from the Greek "heuriskein,"
meaning "to dis
In that sense, a heuristic aims at studying the methods and rules of discovery
[Polya, 1947, pp. 112-113] or assisting in problem solving, which is a process of
but not immediately attainable aim
trying to attain a preconceived
systematically
seen heuristics as procedures to
have
Researchers
[Polya, 1962, 1963]. Operations
to obtain
activities
reduce search in problem-solving
[Tonge, 1961] or a means
a
To
time
within
limited
solutions
[Lin, 1975].
practitioners,
computing
acceptable
The word

cover."

heuristics

are

simple

often

procedures,

good but not necessarily

guided

by

solutions

common

sense,

that

are meant

to

easily

and

to difficult

problems,
optimal
provide
quickly.
... often outlined and
This branch of study is "not very clearly circumscribed,
seldom presented in detail [Polya, 1947]," quick and dirty [Woolsey,
1975a], not
the
mathematical
theorems,
convergence,
optimality,
etiquette (proofs,
abiding by
in analytic purity in the high
that "are conceived
etc.), in contrast to algorithms
in the dark
citadels of academic research, heuristics are midwifed
by expediency
corners of the practitioner's
lair, . . . and are accorded lower status [Glover, 1977]."
For many years heuristics have encountered the hostility, scorn, and condemnation of
tools of the trade, as
them as the poor man's
high academic priests, "treating
with
should
be
that
1977]."
[Eilon,
deep suspicion
regarded
expediencies
as
to
evidenced
this
started
however,
by the growing
change,
picture
Recently,
number of related articles in scholarly reputable journals. The Operations Research
Society of America awarded the 1977 Lancaster Prize to two papers on heuristic
1977; and Karp,
1977]. The
[Cornuejols, Fisher, and Nemhauser,
optimization
Sciences
1980 Innovative Education Award went to
Institute for Decision
American
a paper on heuristic problem
Why then this change?

solving

[Brightman,

1980]. One may naturally wonder:

84

INTERFACES

OPTIMIZATION;HEURISTICS
October

This content downloaded from 125.20.9.248 on Fri, 21 Aug 2015 06:02:37 UTC
All use subject to JSTOR Terms and Conditions

1981

feel that this is caused by a practitioner-researcher-practitioner

slow cyclic
reaction that is now in a stage of rapid acceleration. Practitioners are more pressured
to analyze complex management
by declining productivity
problems with limited
available resources, creating an increased need for good heuristics. Open-minded
We

researchers in academia eventually realize that mathematical

rigor does not guarantee
success of OR/MS systems in practice, and channel some of their efforts towards the
of efficient heuristics that are then eagerly adopted by practitioners.
development
the excellent work of R. Karp [1975a, 1975b, 1976, 1977] accelerated
Undoubtedly,
this reaction in two ways: (1) by proving that many practical combinatorial problems,
cannot be solved efficiently by exact algorithms, whose run
termed NP-complete,
a
with the problem size; and (2) by establishing
ning times increase exponentially
framework

for probabilistic
that elevated them from
analysis of heuristic methods
to a hot topic of sophisticated analysis now respected in the OR/MS

poor relatives
academic

community.

As more

and more researchers take heuristics more seriously

(as M. Fisher
[1980] admits is happening now), improved heuristics and documented performance
thus leading to
analysis will increase their proper and successful use in practice,
additional developments
An
on
increased
heuristic op
reaction).
(cyclic
emphasis
timization should be expected in both academia and businesses over the next several
years.

In this paper, we attempt to provide a unifying framework on the advantages/

of heuristic optimization.
Some of the
uses/misuses,
do's/don'ts,
disadvantages,
issues
been
have
addressed
other
writers
Evans,
1979;
general
[e.g.,
previously
by
and DeWerra,
1980; Woolsey,
1975b; Zanakis,
1977;
and Silver, 1980]. Our emphasis here is to present a
Austin, Nowading,
Thus we
summary of practical guidelines on the use of heuristics for optimization.
prefer the term "heuristic
optimization,"
although the words seem inconsistent.
one may use "heuristic programming"
(more restrictive) or "heuristic
Alternatively,
Ignizio,
Zanakis,

1980; Silver, Vidal,

(more general), which

procedures/algorithms"
lems, such as scheduling.

could include special classes of prob

WHY AND WHEN TO USE HEURISTICS

There are several

instances where

the use of heuristics

is desirable

and advanta

geous:

(1) Inexact or limited data used to estimate model parameters may inherently
contain errors much larger than the "suboptimality"
of a good heuristic. We have
seen monumental
efforts in academia and industry to develop and/or use ILP codes,
for example, to solve large-scale capital budgeting problems with intangible benefit
estimates or risk analysis. In one case [Zanakis, 1977], an existing 0-1 heuristic was
run to produce in 1/10 of CPU time
easily modified within a day and successfully
a
99.7% of the "optimal"
solution to
308x372
capital budgeting problem with
elsewhere with a special
intangible benefits, which had been solved "optimally"
a
six-month developmental
effort.
purpose algorithm after
(2) A simplified model is used, which is already an inaccurate representation of
solution only academic. In both cases 1
the real problem, thus making the "optimal"
exact
and 2, a fast near-optimal
solution makes more sense than a time-consuming
solution to an inexact problem.

INTERFACES
October 1981

85

This content downloaded from 125.20.9.248 on Fri, 21 Aug 2015 06:02:37 UTC
All use subject to JSTOR Terms and Conditions

Small companies may not have

(3) A reliable exact method is not available.
access to a good optimizing code. One of our students, working on an actual problem
for a utility company, found that a classroom heuristic produced a better solution
than the optimization
code the company had been using! And there are problem
solvers in this world that, without questioning,
accept a code's final solution as
optimal.*

but it is computationally
Unattractive due to
(4) An exact method is available,
time and/or storage requirements. Large real-world complex problems may
excessive
prevent an optimizer from finding an optimal or even a feasible solution within a
on the other hand, can produce at least feasible so
reasonable effort. Heuristics,
lutions with minimal

time and storage requirements.

To
the
(5)
improve
of an optimizer, heuristics have been used to
performance
solutions
and/or
provide starting
guide the search and reduce the number of candidate
solutions (e.g., in branch-and-bound).
(6) Repeated need to solve the same problem frequently or on a real-time basis,
which may translate a heuristic's
speed into significant computing savings.
A
heuristic
be
solution
(7)
may
good enough for a manager if it produces results
better

than those currently realized.

are simple so users can understand them, gain confidence
in the
(8) Heuristics
more
and
to
be
the
therefore
recommended
approach,
likely
changes.
implement
(9) As a learning device, heuristics may help in gaining insight into a complex
and solved more pragmatically.
problem, which then may be modeled
limitations
(10) Other resource
and
etc.) may
availability
knowledge,

(e.g., project time and budget,

enforce use of a heuristic.

manpower

Most of the above reasons have been advocated by practice-oriented

persons and
are epigramatically
summarized elsewhere: "Managers want to improve their current
operation as cheaply and quickly as possible, care little about an optimal solution to a
problem with usually inexact data, and will not accept a new solution they do not
understand
and Silver, 1980]." As T. Levitt [1978]
[Zanakis, Austin, Nowading,
is the only way to manage
noted, simplicity
anything, even in today's complex
world.

We are not advocating

that optimizing methods be dispensed with; indeed, if
an
one can effectively
apply
optimization
algorithm to solve a problem, then by all
means do. In large-scale applications,
such as facility location for instance, the
a
or
two
difference of
from optimality may mean several hundred thousand
percent
dollars to the decision maker. For such capital-intensive
decisions, more care would
most
and model building;
therefore an optimizing
likely go into data collection
technique would
reasons outlined

be appropriate. However,
for many real-world
above often suggest a heuristic method.

applications,

the

FEATURES OF GOOD HEURISTICS

In general, and without regard to a specific problem, a good heuristic
have the following properties or features:
(1) Simplicity, which facilitates user understanding and acceptance.
core storage requirements.
(2) Reasonable

should

(3) Speed, i.e., reasonable computing times that do not grow polynomially
even exponentially
as problem size increases.
*Ed. Note:

I observed

this

10 years

ago

inManagement

86

Science,

1971,

Vol.

17, No.

or

7, p. 500.

INTERFACESOctober 1981

This content downloaded from 125.20.9.248 on Fri, 21 Aug 2015 06:02:37 UTC
All use subject to JSTOR Terms and Conditions

(4) Accuracy,
i.e., small average and mean square errors; these will increase
chances
the
of a good (poor) solution. The degree of acceptable accuracy
(decrease)
should be dictated by the problem and the user.
should obtain good solutions,
in reasonable
i.e., the method
(5) Robustness,
a
for
wide
times,
variety of problems and not be too sensitive to changes in paramet
ers. This is not true of many branch-and-bound methods where a small change in a
parameter may double or triple running times.
?
a formidable
feasible
(6) Accept multiple starting points, not necessarily
task in large and severely constrained problems.
(7) Produce multiple solutions
(ideally in a single run) by properly selecting
input parameters or perturbing the final solution. This allows the user to select the
result that is most accurate or "satisficing"
(in case of unquantifiable
intangibles).
A new perturbation heuristic, added to several existing 0-1 LP heuristics,
captured
about 65% of possible improvement to optimality,
thus increasing average accuracy
to 99.8%

[Zanakis and Gillenwater,

(8) Good stopping criteria that
These are in general
"stagnation."
provement per additional computing
(9) Statistical estimation of the
next
section and appendix).
(see
(10) Interactive

ability with

1981].
take advantage of search "learning"
and avoid
some ratio measure of objective function im
effort.
true optimum from a series of iterative solutions

the analyst and decision maker.

An ideal heuristic is one that has all of the above properties. It should be noted
that problem-dependent
that take advantage of the special structure of a
heuristics,
Evans
and
Cullen,
1977], are more efficient than general mathemati
problem [e.g.,
cal programming heuristics, but their use is limited only to a specific class of prob
lems. A classical example
is scheduling heuristics vs integer linear programming
heuristics applied to scheduling problems.

HOW TO USE HEURISTICS

Many people may argue that proper use of heuristics is more of an art than a
science. It is certainly based on common sense, but some form of analytical work is
needed to increase confidence
in the heuristic results (current and future). A few
are
general guidelines
presented below:
and
(1) Use heuristics carefully and cautiously; understand their capabilities
limitations. Don't fall in love with them (a tendency of many analysts towards their
intellectual baby).
(2) Decide whether a heuristic solution on the total problem is preferable to an
exact solution of heuristically obtained subproblem(s),
i.e., solution space reduction
vs problem simplification
and structural
(decomposition,
partitioning,
aggregation,
approximation). A more detailed discussion of these heuristic strategies is provided
by Evans [1979]. Random, biased, and improvement
sampling for generating so
lutions to large combinatorial
is
heuristic
another
problems
approach [Mabert and
1977].
Whybark,
the model before solution; i.e., check data accuracy, rearrange (or
(3) "Clean"
eliminate redundant) variables and constraints, etc. We have often noticed improved
performance of heuristics when such redundancy was eliminated.
(4) Be aware of the pitfalls of the common practice of rounding continuous
1975].
optimal solutions in discrete-variable
problems [Glover and Sommer,

INTERFACES
October 1981

87

This content downloaded from 125.20.9.248 on Fri, 21 Aug 2015 06:02:37 UTC
All use subject to JSTOR Terms and Conditions

such as CPU time, number

(5) Establish appropriate measures of performance,
of iterations, percent of times optimal, average (percent) error, mean-squared
error,
error (not necessarily correlated with average error), etc.
maximum
(6) Solve a sufficient number of literature and large-scale randomly generated
test problems of the type for which
the method was designed. The challenging
difficulty of the former is usually structure, while for the latter it is size. Literature
test problems are hard-to-solve
benchmark problems, usually with known optimal
thus of small size. Of great importance is the generation of "realistic"
solutions ?
large-size problems with known optimal solutions (which cannot be determined with
an exact method). This has been achieved for linear programming
[Chames et al.,
more
and
and
nonlinear
Suzuki,
1965],
[Rosen
recently for
1974],
programming
test problems
1976; Rardin and Lin,
[Fleisher and Meyer,
integer programming
1980].
solution results statistically. Use analysis of variance
[Zanakis,
(7) Validate
and
1977]
experimental design approaches [Lin and Rardin, 1979] to understand the
i.e., when to
relationships between problem characteristics and method performance,
use or not use a certain heuristic.
(8) Predict or bound the performance of a heuristic by using one or more of the
approaches:
to establish upper and lower bounds on the op
relaxation
(a) Problem
some con
some
timum by relaxing
problem requirement, e.g., integral variables,
straints), etc.
the maximum possible error
(b) Worst-case
study to establish analytically
a
a
on
certain class of problems [Christofides,
heuristic
of
1976; Cornuejols et al.,
following

1977; Fisher,

1980].
analysis, where one assumes a probability distribution of
statistical properties of a heuristic
[Karp, 1976, 1977].
functions can provide tractable results, which is a limitation of

(c) Probabilistic
data to establish

problem
Only simple density
this approach.
(d) Statistical
true (unknown) optimal

inference to determine point and interval estimates of the

value based on a series (sample) of iterative solutions. The
idea of point estimation for large combinatorial
problems was proposed earlier by
others
later
extended
Swart [1967] and
1977; Golden and Alt,
[e.g., Golden,
by
and
Zanakis
Gillenwater,
1981]. These procedures are
1977;
1979; Dannenbring,
to
based on the statistical property of the smallest (or largest) sample observation
the
when
or a three-parameter Weibull
follow an extreme-value
distribution,
sample
size gets

widely,

large. Because of its simplicity and the fact that it has not been publicized
this estimation procedure is summarized in an appendix.

CONCLUSIONS
The need for and usage of good heuristics in both academia and businesses will
continue increasing fast. When confronted with real world problems, a researcher in
of seeing his product,
academia experiences at least once the painful disappointment
a theoretically sound and mathematically
procedure, not used by its
"respectable"
ultimate user. This has encouraged researchers to develop new improved heuristics
thus spreading further their usage in
and rigorously
evaluate their performance,
practice, where heuristics have been advocated for a long time. More work is needed
case [Zanakis,
in the area of heuristic optimization, particularly in the multiobjective
to
world
meet
real
the
fail
the
codes
available
where
problem needs.
1981],
seriously

88

INTERFACESOctober 1981

This content downloaded from 125.20.9.248 on Fri, 21 Aug 2015 06:02:37 UTC
All use subject to JSTOR Terms and Conditions

References
Problem
H.,
1980, "Teaching
Quantitative
by General
Techniques
Solving Heuristics,"
AIDS 12th National
pp. 116-118.
Proceedings
Conference,
of Test Problems
Com
for Linear Programming
Chames,
A., et al.,
1974, "On Generation
Codes,"
munications
Vol.
17, No.
10, pp. 583-586.
oftheACM,
of a New Heuristic
for the Traveling-Salesman
Christofides,
N.,
1976, "Worst-Case
Problem,"
Analysis
Brightman,

MS
Cooke,

Report No. 388, Carnegie-Mellon

P. 1979, "Statistical
Inference
367-374.

University,
February.
for Bounds
of Random

Biometrika

Variables,"

Vol.

66,

pp.

to
G. L.,
of Bank Accounts
G., Fisher, M. L., and Nemhauser,
1977, 1979, "Location
Float: An Analytic
Science
Optimize
Study of Exact and Approximate
Algorithms,"
Management
Vol. 23, pp. 789-810
Vol. 25, pp. 808-809).
Science
(Errata inManagement
D. G.,
Solution Values
for Estimating
for Large Combinato
1977, "Procedures
Dannenbring,
Optimal
rial Problems,"
Science Vol. 23, pp. 1273-1283.
Management
Cornuejols,

S., 1977, "More Against

J. R.,
1979, "Toward

Eilon,
Evans,

Optimization,"
a Framework

Omega Vol. 5, pp. 627-633.

for Heuristic
Optimization,"

Proceedings

1979, pp. 284-292.

Conference,
Spring
F. H.,
J. R. and Cullen,
1977, "The Segregated
Evans,
Storage
AIIE Transactions
Effective
Vol. 9, pp. 409-413.
Heuristic,"

Problem:

Fisher, M. L.,
1-17.

Management

1980, "Worst-Case

Analysis

of Heuristic

Algorithms,"

Some

Annual

AIIE

and an

Properties

Science

Vol.

26, pp.

J. M.

and Meyer,
R. R.,
Sufficient
Conditions
for Integer Program
1976, "New
Optimality
and Their Application,"
Science Technical
of Wiscon
Computer
Report No. 278, University
October.
sin, Madison,
for Combinational
Problems: An
Garey, M. R. and Johnson, D. S., 1976, "Approximation
Algorithms
Fleisher,

ming

Annotated

in Algorithms
and Complexity,
Bibliography,"
F. and Sommer, D. C,
of Rounding
1975, "Pitfalls
Vol. 6, pp. 211-220.
Sciences
lems," Decision
for Integer Programming
F., 1977, "Heuristics
Glover,
Using
Glover,

Vol.

8, pp.

J. F. Traub, ed., Academic

Press.
in Discrete Management
Decision
Surrogate

Constraints,"

Decision

Prob

Sciences

156-166.

B. L., 1977, "A Statistical Approach

to the TSP," Networks
B. L. and Alt, F. B., 1979, "Interval Estimation
of a Global
Naval Research
Problems,"
pp. 69-77.
Logistics
Quarterly,

Vol.

Golden,
Golden,

7, pp. 209-225.
for Large Combinatorial

Optimum

A Venture
Problems:
into a New Dimension,"
Journal
1980, "Solving
Large-Scale
Research
Society Vol. 31, pp. 217-225.
Vol.
of Combinational
Networks
1975a, "On the Computational
Problems,"
Karp, R.,
Complexity
pp. 45-68.
Ignizio,

J. P.,

of

Operational

5,

Solution of Hard Combinatorial

6th
R., 1975b, "The Fast Approximate
Problems,"
Proceedings
SE Conference
Combinatorics,
pp. 15-31.
Graph Theory, & Computing,
in Algorithms
of Some Combinatorial
Search Algorithms,"
Karp, R., 1976, "The Probabilistic
Analysis
and Complexity
Press.
(J. F. Traub, ed.), Academic

Karp,

R. M.,
of Partitioning
for the Traveling-Salesman
1977, "Probabilistic
Analysis
Algorithms
Problem
in the Plane," Mathematics
Research
Vol. 2, pp. 209-224.
of Operations
View of Management
December
Levitt, T., 1978, "A Heretical
Fortune,
'Science',"
18, pp. 50-52.

Karp,

S., 1975, "Heuristic

B. W. and Rardin,

Lin,
Lin,

Programming
R. L.,
1979,

as an Aid

Vol. 5, pp. 33-43.

Networks
Design,"
for Statistical
of
Experimental
Design
Comparison
Science Vol. 25, pp. 1258-1270.
Management
as a Solution Methodology,"
Sciences
Decision
1977, "Sampling

Integer Programming
Algorithms,"
V. A. and Whybark,
D. C,
Mabert,
Vol. 8, pp. 167-177.
to Solve It, Princeton
1947, How
Polya, G.,
Polya,

G.,

1962, Mathematical

to Network

"Controlled

Discovery,

Press, Princeton.
University
Vol.
New York.
I,Wiley,
New York.
Vol.
II, Wiley,

1965, Mathematical
Polya, G.,
Discovery,
Test Problem Generator,"
Rardin, R. L. and Lin, B. W.,
1980, "The RIP Random
Integer Programming
School of Industrial & Systems Engineering,
Institute of Technology,
June.
Georgia
J. B. and Suzuki,
of Nonlinear
Test Problems,"
Com
Rosen,
S., 1965, "Construction
Programming
munications

oftheACM

Vol.

8, No.

2, p.

113.

INTERFACESOctober 1981 89

This content downloaded from 125.20.9.248 on Fri, 21 Aug 2015 06:02:37 UTC
All use subject to JSTOR Terms and Conditions

E. A., Vidal, R. V. V., and De Werra,

1980, "An Introduction
D.,
Vol. 5, pp. 153-162.
Journal of Operational
Research
European
to the Redistricting
Value Estimation
Swart, W. W.,
1967, "Optimal
Applied
MSIE Thesis, Georgia
Institute of Technology.

Silver,

to Heuristic

Methods,"

Problem,"

unpublished

in Management
F. M.,
Science,"
1961, "The Use of Heuristic
Management
Tonge,
Programming
Vol. 7, pp. 231-237.
Science
A Quick
R. E. D. and Swanson, H. S., 1975a, Operations
Research for Immediate Application:
Woolsey,
and Dirty Manual,
Harper & Row, New York.
to Do Integer Programming
R. E. D.,
1975b, "How
H. M. Salkin, ed., Addison-Wesley,
13.
Chap.
S. H.,
0-1 Linear Programming:
Zanakis,
1977, "Heuristic
Vol. 24, pp. 91-104.
Science
Methods,"
Management

Woolsey,

in the Real World,"

in Integer

Program

ming,

An

Experimental

Comparison

S. H., 1979a, "A Simulation

of the Three-Parameter
Study of Some Simple Estimators
and Simulation
Vol. 9, pp. 101-116.
Journal
Distribution,"
of Statistical
Computation
S. H., 1979b, "Extended
for the Three-Parameter
Pattern Search with Transformations
Zanakis,
MLE Problem," Management
Vol. 25, pp. 1149-1161.
Science

Zanakis,

Zanakis,

of Three
Weibull
Weibull

to Implementing
1980, "From Teaching
D., and Silver, E.,
of Translation,"
Vol.
10, No. 6, pp. 103-110.
Interfaces
to a
Method
with an Application
Large-Scale
Integer Goal Programming
in Organizations:
Multicriteria
Decision
Location/Allocation
Problem,"
Multiple
Making,

S. H.,

Austin,

L., Nowading,
Problems

Inventory Management:
S. H.,
Zanakis,
1981, "A
Facility
Agents
Zanakis,

with Multiple
Criteria,
and Gillenwater,

S. H.,

Working

Paper, West

Virginia

J. Morse,
J.,

1981,

College

ed. Springer-Verlag.
Improved Heuristics
of Graduate

for

Integer

and 0-1

Programming,

Studies.

90 INTERFACESOctober 1981

This content downloaded from 125.20.9.248 on Fri, 21 Aug 2015 06:02:37 UTC
All use subject to JSTOR Terms and Conditions

APPENDIX
Let Z, ^ Z2 ^ .... ^ Zn be an ordered random sample of iterative solutions
(For a maximization
(objective function values) to a large minimization
problem.
..
i=l, 2,
.,n.) A point estimate of the
problem,
simply replace Z? by -Zn+1_f,
unknown true optimum can be obtained by one of the following simple estimators:

Z = 2Z,-Z2 (1)
used by Dannenbring

[1977] and Golden

[1977], or

Z = (Z, Z? - Z|)/(ZX + Z? - 2Z2)

used by Zanakis

[1979a]

and Zanakis

Z* =

and Gillenwater

2Zj -(e-1)

[1981],

(2)

or

??=i Z.?? (3)

by Cooke [1979]. Note that these estimators do not require any restrictive
assumption and apply to any distribution with a location or threshold parameter.
From a storage requirement viewpoint
their order of preference
is Z, Z, Z* but
proposed

their accuracy (MSE) is in the reverse order.

number of observations
considered by each
The following very simple and robust
unknown true optimum value, Zopt=a9 was

This

is expected
estimator.

100(l-e~w)%

because

confidence

by Golden

proposed
< a < Zj
6, (4)
Zj

where

of the different
interval on the

and Alt

[1979]:

=Z[o.63nl

+ i ""*

(5)

=
Here, [
] denotes rounding down and af Z (Z or Z* will be more accurate).
This requires the (empirically
that the random sample came
justified) assumption
from a three-parameter Weibull
cumulative distribution

F(Z)

= 1exp{- [(Z a)lb]c},

Z>a,

(6)

parameters a (location), b (scale), and c (shape). Golden and Alt [1979] rec
ommend refining this confidence
interval by using maximum
likelihood estimates
(MLE). The added accuracy, however, will not justify the incremental effort be
cause: (1) MLE's do not have a closed form, thus requiring an iterative procedure on
a computer; (2) it will require the additional estimation of the shape parameter c,
which is the most difficult of the three parameters to estimate; (3) estimator b is a

with

scale estimator, even better than MLE's when c is small

very good simple Weibull
=
Z*, or even Z, instead of Z, will
[Zanakis, 1979a and 1979b]; and (4) using ?
the simple confidence
interval (4)
interval. Therefore,
improve the above confidence
should suffice for most simple heuristic optimization methods.

INTERFACESOctober 1981

91

This content downloaded from 125.20.9.248 on Fri, 21 Aug 2015 06:02:37 UTC
All use subject to JSTOR Terms and Conditions