An Introduction to Mathematical Modeling
5/5
()
About this ebook
The lively and accessible text requires only minimal scientific background. Designed for senior college or beginning graduate-level students, it assumes only elementary calculus and basic probability theory for the first part, and ordinary differential equations and continuous probability for the second section. All problems require students to study and create models, encouraging their active participation rather than a mechanical approach.
Beyond the classroom, this volume will prove interesting and rewarding to anyone concerned with the development of mathematical models or the application of modeling to problem solving in a wide array of applications.
Read more from Edward A. Bender
A Short Course in Discrete Mathematics Rating: 3 out of 5 stars3/5
Related to An Introduction to Mathematical Modeling
Related ebooks
Concepts of Mathematical Modeling Rating: 4 out of 5 stars4/5Advanced Mathematics for Engineers and Scientists Rating: 4 out of 5 stars4/5Elementary Matrix Algebra Rating: 3 out of 5 stars3/5Vector Spaces and Matrices Rating: 0 out of 5 stars0 ratingsIntroduction to Matrices and Vectors Rating: 0 out of 5 stars0 ratingsIntroduction to Proof in Abstract Mathematics Rating: 5 out of 5 stars5/5Introduction to Analysis Rating: 4 out of 5 stars4/5Analysis of Numerical Methods Rating: 3 out of 5 stars3/5Nonlinear Programming: Analysis and Methods Rating: 5 out of 5 stars5/5Sets, Sequences and Mappings: The Basic Concepts of Analysis Rating: 0 out of 5 stars0 ratingsMatrices and Transformations Rating: 4 out of 5 stars4/5Introduction to the Foundations of Mathematics: Second Edition Rating: 4 out of 5 stars4/5Journey into Mathematics: An Introduction to Proofs Rating: 4 out of 5 stars4/5Complex Analysis Rating: 3 out of 5 stars3/5Introduction to Vector and Tensor Analysis Rating: 4 out of 5 stars4/5Introduction to Real Analysis Rating: 3 out of 5 stars3/5The Skeleton Key of Mathematics: A Simple Account of Complex Algebraic Theories Rating: 0 out of 5 stars0 ratingsHow to Solve Mathematical Problems Rating: 4 out of 5 stars4/5Shape Theory: Categorical Methods of Approximation Rating: 0 out of 5 stars0 ratingsElements of Abstract Algebra Rating: 4 out of 5 stars4/5Applied Multivariate Analysis: Using Bayesian and Frequentist Methods of Inference, Second Edition Rating: 0 out of 5 stars0 ratingsIntroduction to Mathematical Thinking: The Formation of Concepts in Modern Mathematics Rating: 4 out of 5 stars4/5Introduction to Linear Algebra and Differential Equations Rating: 3 out of 5 stars3/5Linear Mathematics: A Practical Approach Rating: 5 out of 5 stars5/5Mathematics for the Nonmathematician Rating: 4 out of 5 stars4/5Finite-Dimensional Vector Spaces: Second Edition Rating: 0 out of 5 stars0 ratingsBasic Abstract Algebra: For Graduate Students and Advanced Undergraduates Rating: 4 out of 5 stars4/5Splines and Variational Methods Rating: 5 out of 5 stars5/5
Mathematics For You
The Little Book of Mathematical Principles, Theories & Things Rating: 3 out of 5 stars3/5Basic Math & Pre-Algebra For Dummies Rating: 4 out of 5 stars4/5Quantum Physics for Beginners Rating: 4 out of 5 stars4/5Standard Deviations: Flawed Assumptions, Tortured Data, and Other Ways to Lie with Statistics Rating: 4 out of 5 stars4/5Mental Math Secrets - How To Be a Human Calculator Rating: 5 out of 5 stars5/5What If?: Serious Scientific Answers to Absurd Hypothetical Questions Rating: 5 out of 5 stars5/5My Best Mathematical and Logic Puzzles Rating: 4 out of 5 stars4/5Calculus Made Easy Rating: 4 out of 5 stars4/5Algebra - The Very Basics Rating: 5 out of 5 stars5/5Algebra II For Dummies Rating: 3 out of 5 stars3/5The Golden Ratio: The Divine Beauty of Mathematics Rating: 5 out of 5 stars5/5Real Estate by the Numbers: A Complete Reference Guide to Deal Analysis Rating: 0 out of 5 stars0 ratingsThe Everything Guide to Algebra: A Step-by-Step Guide to the Basics of Algebra - in Plain English! Rating: 4 out of 5 stars4/5Limitless Mind: Learn, Lead, and Live Without Barriers Rating: 4 out of 5 stars4/5Calculus For Dummies Rating: 4 out of 5 stars4/5Algebra I Workbook For Dummies Rating: 3 out of 5 stars3/5Mental Math: Tricks To Become A Human Calculator Rating: 5 out of 5 stars5/5The Math of Life and Death: 7 Mathematical Principles That Shape Our Lives Rating: 4 out of 5 stars4/5The Everything Everyday Math Book: From Tipping to Taxes, All the Real-World, Everyday Math Skills You Need Rating: 5 out of 5 stars5/5ACT Math & Science Prep: Includes 500+ Practice Questions Rating: 3 out of 5 stars3/5The New York Times Book of Mathematics: More Than 100 Years of Writing by the Numbers Rating: 0 out of 5 stars0 ratingsPre-Calculus For Dummies Rating: 5 out of 5 stars5/5How to Calculate Quickly: Full Course in Speed Arithmetic Rating: 4 out of 5 stars4/5Relativity: The special and the general theory Rating: 5 out of 5 stars5/5Painless Algebra Rating: 0 out of 5 stars0 ratingsHow Math Explains the World: A Guide to the Power of Numbers, from Car Repair to Modern Physics Rating: 3 out of 5 stars3/5How to Solve It: A New Aspect of Mathematical Method Rating: 4 out of 5 stars4/5
Reviews for An Introduction to Mathematical Modeling
1 rating0 reviews
Book preview
An Introduction to Mathematical Modeling - Edward A. Bender
INDEX
CHAPTER 1
WHAT IS MODELING
1.1 MODELS AND REALITY
The theoretical and scientific study of a situation centers around a model, that is, something that mimics relevant features of the situation being studied. For example, a road map, a geological map, and a plant collection are all models that mimic different aspects of a portion of the earth’s surface.
The ultimate test of a model is how well it performs when it is applied to the problems it was designed to handle. (You cannot reasonably criticize a geological map if a major highway is not marked on it; however, this would be a serious deficiency in a road map.) When a model is used, it may lead to incorrect predictions. The model is often modified, frequently discarded, and sometimes used anyway because it is better than nothing. This is the way science develops.
Here we are concerned exclusively with mathematical models, that is, models that mimic reality by using the language of mathematics. Whenever we use model
without a modifier, we mean mathematical model.
What makes mathematical models useful? If we speak in mathematics,
then
We must formulate our ideas precisely and so are less likely to let implicit assumptions slip by.
We have a concise language
which encourages manipulation.
We have a large number of potentially useful theorems available.
We have high speed computers available for carrying out calculations.
There is a trade-off between items 3 and 4: Theory is useful for drawing general conclusions from simple models, and computers are useful for drawing specific conclusions from complicated models. Since the thought habits needed in formulating models are quite similar in the two cases, it matters little what sort of models we use; consequently, I have felt free to neglect computer based models purely for personal pedagogical reasons. There are some references to a computer in Section 5.2 where Monte Carlo simulation is discussed and, to a lesser extent, in Section 8.2 where numerical solutions to differential equations are discussed.
Mathematics and physical science each had important effects on the development of the other. Mathematics is starting to play a greater role in the development of the life and social sciences, and these sciences are starting to influence the development of mathematics. This sort of interaction is extremely important if the proper mathematical tools are going to be developed for the various sciences. S. Bochner (1966) discusses the hand-in-hand development of mathematics and physical science. Some people feel that there is something deeper going on than simply an interaction leading to the formulation of appropriate mathematical and physical concepts. E. P. Wigner (1960) discusses this.
1.2. PROPERTIES OF MODELS
We begin with a definition based on the previous discussion: A mathematical model is an abstract, simplified, mathematical construct related to a part of reality and created for a particular purpose. Since a dozen different people are likely to come up with a dozen different definitions, don’t take this one too seriously; rather, think of it as a crude starting point around which to build your own understanding of mathematical modeling.
We now have a problem: To fully appreciate the general discussion in the next two sections you should look at some concrete examples like those in Sections 1.4 and 1.5; however, you will need some abstract background to appreciate the examples fully. I suggest reading the remainder of the chapter through quickly and then coming back to this point and rereading more carefully.
As far as a model is concerned, the world can be divided into three parts:
Things whose effects are neglected.
Things that affect the model but whose behavior the model is not designed to study.
Things the model is designed to study the behavior of.
The model completely ignores item 1. The constants, functions, and so on, that appear in item 2 are external and are referred to as exogenous variables (also called parameters, input, or independent variables). The things the model seeks to explain are endogenous variables (also called output or dependent variables). The exogenous-endogenous terminology is used in some areas of modeling. The input-output terminology is used in areas of modeling where the model is viewed as a box into which we feed information and obtain information from. The parameter-independent-dependent terminology is the standard mathematical usage.
Suppose we are hired by a firm to determine what the level of production should be to maximize profits. We would construct a model that enables us to express profits (the dependent variable) in terms of the level of production, the market situation, and whatever else we think is relevant (the independent variables). Next we would measure all the independent variables except the level of production and use the model to determine which value of the level of production gives the greatest profit.
Now let’s look at things from the point of view of an economist who is seeking to explain the amount of goods firms produce. A two-part model could be constructed: Firms seek to maximize profits, and profits can be determined as sketched in the previous paragraph. In this model profits become an internal variable (of no interest except for the machinations of the model), and level of production changes from an independent to a dependent variable.
These three categories (neglected, input, and output) are important in modeling. If the wrong things are neglected, the model will be no good. If too much is taken into consideration, the resulting model will be hopelessly complex and probably require incredible amounts of data. Sometimes, in desperation a modeler neglects things not because he thinks they are unimportant, but because he cannot handle them and hopes that neglecting them will not invalidate the conclusions. A. Jensen (1966) discusses the development of a model for safety-at-sea problems. The main difficulty in formulating the model was to determine what types of encounters between ships were dangerous, that is, to separate items 1 and 2. He found this to be hard even with the aid of nautical experts. (If you want to know the answer, you’ll have to read the article.)
Proper choice of dependent variables (i.e., output) is essential; we must seek to explain the things we can explain. Often this choice is relatively clear, as in the example involving the economist who wished to explain the level of production of a firm. Sometimes we need to be careful; for example, we could explain profits in terms of level of production, but not conversely as we might naively try to do, since we were asked to determine the best level of production.
Since different models make different types of simplifying assumptions, there is usually no single best model for describing a situation. R. Levins (1968, p. 7) observed that it is not possible to maximize simultaneously generality, realism, and precision.
In the social sciences one is often content with a statement that something will increase; precision has been sacrificed for realism and (hopefully) generality. Simulation models usually try for precision and realism but sacrifice generality. These three trade-offs should become clearer after you have studied some actual models.
Definitions of the variables and their interrelations constitute the assumptions of the model. We then use the model to draw conclusions (i.e., to make predictions). This is a deductive process: If the assumptions are true, the conclusions must also be true. Hence a false prediction implies that the model is wrong in some respect. Unfortunately things are usually not this clear-cut. We know our model is only an approximation, so we cannot expect perfect predictions. How can we judge a model in this case?
A conclusion derived from a crude model is not very believable, especially if other models make contrary predictions. A result is robust if it can be derived from a variety of different models of the same situation, or from a rather general model. A prediction that depends on very special assumptions for its validity is fragile. The cruder the model, the less believable its fragile predictions.
You may notice that we have talked about conclusions, not explanations. Can a model provide explanations? This is a somewhat philosophical question, and different people have different notions of what constitutes an explanation. Let us grant that, in some sense, models can provide explanations. A decision about the validity of a model is usually based on the accuracy of its predictions. Unfortunately, two different models may make the same predictions but offer different explanations. How can this be?
We can think of the situation we are modeling as being a black box
which outputs something for every input. (Something
can be no output.) A model makes correct predictions if it outputs the model equivalent of the black box output whenever the model equivalent of the black box input is fed in. The mechanism is irrelevant when dealing with predictions, but the nature of the mechanism is the heart of an explanation. Although there is usually a situation in which two different models lead to different predictions, we may not be able to determine which prediction is correct. For example, a model of a politician can be constructed by assuming that his behavior is (1) motivated by concern for his fellow man or (2) motivated by a desire for public office. In many situations these two models lead to identical or very similar predictions. It may be difficult to make contradictory predictions that can be checked. Another example for those familiar with simple circuits is the mathematical equivalence between perfect springs and perfect LC circuits. Although the underlying mathematics is identical, no one would seriously suggest that Hooke’s law for springs explains
the circuit’s behavior.
We have been talking about an ideal modeler. When any of us approaches a problem, we do so in a limited, biased fashion. The more open-minded, communicative, and creative we can be, the better our model is likely to be. The following poem illustrates the problems that can arise.
The Blind Men and the Elephant
It was six men of Indostan
To learning much inclined,
Who went to see the Elephant
(Though all of them were blind),
That each by observation
Might satisfy his mind.
The First approached the Elephant,
And happening to fall
Against his broad and sturdy side,
At once began to bawl:
"God bless! but the Elephant
Is very like a wall!"
The Second, feeling of the tusk,
Cried, "Ho! what have we here
So very round and smooth and sharp?
To me ’tis mighty clear
This wonder of an Elephant
Is very like a spear ! "
The third approached the animal,
And happening to take
The squirming trunk within his hands,
Thus boldly up and spake:
I see,
quoth he, "the Elephant
Is very like a Snake!"
The Fourth reached out an eager hand,
And felt about the knee.
"What most this wondrous beast is like
Is mighty plain," quoth he;
" ’Tis clear enough the Elephant
Is very like a tree!"
The Fifth who chanced to touch the ear,
Said: "E’en the blindest man
Can tell what this resembles most;
Deny the fact who can,
This marvel of an Elephant
Is very like a fan!"
The Sixth no sooner had begun
About the beast to grope,
Than, seizing on the swinging tail
That fell within his scope,
I see,
quoth he, "the Elephant
Is very like a rope!"
And so these men of Indostan
Disputed loud and long,
Each in his own opinion
Exceeding stiff and strong.
Though each was partly in the right
And all were in the wrong!
John Godfrey Saxe (1816–1887)
Reprinted in Engineering Concepts
Curriculum Project (1971)
1.3. BUILDING A MODEL
Model building involves imagination and skill. Giving rules for doing it is like listing rules for being an artist; at best this provides a framework around which to build skills and develop imagination. It may be impossible to teach imagination. It won’t try, but I hope this book provides an opportunity for your skills and imagination to grow. With these warnings, I present an outline of the modeling process.
Formulate the Problem. What is it that you wish to know? The nature of the model you choose depends very much on what you want it to do.
Outline the Model. At this stage you must separate the various parts of the universe into unimportant, exogenous, and endogenous. The interrelations among the variables must also be specified.
Is It Useful? Now stand back and look at what you have. Can you obtain the needed data and then use it in the model to make the predictions you want? If the answer is no, then you must reformulate the model (step 2) and perhaps even the problem (step 1). Note that useful
does not mean reasonable or accurate; they come in step 4. It means: If the model fits the situation, will we be able to use it?
Test the Model. Use the model to make predictions that can be checked against data or common sense. It is not advisable to rely entirely on common sense, because it may well be wrong. Start out with easypredictions—don’ t waste time on involved calculations with a model that may be no good. If these predictions are bad and there are no mathematical errors, return to step 2 or step 1. If these predictions are acceptable, they should give you some feeling for the accuracy and range of applicability of the model. If they are less accurate than you anticipated, it is a good idea to try to understand why, since this may uncover implicit or false assumptions.
At this point the model is ready to be used. Don’t go too far; it is dangerous to apply the model blindly to problems that differ greatly from those on which it was tested. Every application should be viewed as a test of the model.
You may not be able to carry out step 2 immediately, because it is not clear what factors can be neglected. Furthermore, it may not be clear how accurately the exogenous variables need to be determined. A common practice is to begin with a crude model and rough data estimates in order to see which factors need to be considered in the model and how accurately the exogenous variables must be determined.
Some models may require no data. If a model makes the same prediction regardless of the data, we are not getting something for nothing because this prediction is based on the assumptions of the model. To some extent, the distinction between data and assumptions is artificial. In an extreme case, a model may be so specialized that its data are all built into the assumptions.
Sometimes step 4 may be practically impossible to carry out. For example, how can we test a model of nuclear war? What do we do if we have two models of a nuclear war and they make different predictions? This can easily happen in fields of study that lack the precisely formulated laws found in the physical sciences. At this point experience is essential—not experience in mathematics but experience in the field being modeled. Even if predictions can be tested, the testing may be expensive to carry out and may require training in a particular field of experimental science. Since the absence of experimental verification leaves the modeling process incomplete, I have given test results whenever I have been able to obtain them.
1.4. AN EXAMPLE
We discuss models for the long term growth of a population in order to illustrate some of the ideas of the two previous sections. We want to predict how a population will grow numerically over a few generations. This is the problem (step 1 in Section 1.3).
Let the exogenous (independent) variables be the net reproduction rate r per individual, the time t, and the size of the population at t = 0. The net reproduction rate is the birth rate minus the death rate. In other words, it is the fractional rate of change of the population size: r = (dN/dt)/N. There is only one endogenous (dependent) variable, the size of the population at time t, which we denote by N(t). We also refer to r as the net growth rate.
To obtain a simple model, we ignore time lag effects; that is, we assume that only the present value of N and its derivatives are relevant in determining the future values of N. (This will lead to a differential equation.) If the fraction of the population that is of reproductive age varies with t, this can be a very poor assumption. Let’s also assume that the net reproduction rate r is a constant. This gives us a rather crude model with the basic relationship
(1)
The model would certainly be useful if it fits the real world (step 3). The solution of (1) is N(t) = N(0)ert. Unless r = 0, the population will eventually either die out (r negative) or grow to fill the universe (r positive). Reasonable behavior of the population size is a very fragile prediction of the model. This casts serious doubt on the validity of using a constant net reproduction rate for predicting long term growth. This approach to a model illustrates an important point: Study the behavior of your model in limiting cases (in this case as time gets very long, i.e., as t → ∞).
Our test of the model (step 4) for long term growth indicates that it must be rejected; however, it may be useful for short term predictions. Unfortunately, we specifically asked for long term predictions.
Clearly the growth rate of a population will depend on the size of the population because of such effects as exhaustion of the food supply. If the population becomes very large, we can expect the death rate to exceed the birth rate. Let’s translate this into mathematics. We replace the net reproduction rate r in (1) by r(N) which is a strictly decreasing function of N for large N and becomes negative when N is very large. Thus
(2)
We’ve now redone step 2. The model is less useful than the previous one, because obtaining the exact form of r(N) will be hard, perhaps even impossible. However, rough estimates can be obtained, so let’s see what can be done with them. On to step 4. It can be shown that N(t) approaches N0, the solution of r(N0) = 0, as time passes. This is a robust prediction, since we made very few assumptions concerning the nature of the function r(N). Because the model was constructed to predict an upper limit for the size of a population, it is not surprising that it does so.
The cycle of steps 4, 2, and 3 can be repeated, since the model described by (2) has many drawbacks. For one thing, the population can only move closer to N0 in the future. A real population often overshoots the steady state size N0, and even steady state populations fluctuate slightly in size because of the somewhat random nature of births and deaths. One way to eliminate the first objection is to introduce time lags. For example, if the death rate m is not age dependent and the birth rate b changes from zero to a constant at age p, we could replace (1) by
(3)
The parameter p is called a time lag. Of course, we could make m and b functions of N(t), N(t − p), or some weighted average of N on the interval [t − p, t]. To allow for random fluctuations we must replace our deterministic model by a random one.
Another drawback is the assumption that it makes sense to talk about r(N). If the age or sex ratios in a population are changing, this may be nonsense. To overcome this objection it is necessary to split the population into subpopulations based on age and sex. Demographic models are designed in this way: In a typical model time is broken up into discrete units such as 5 year periods, men are ignored, and women are divided into age classes separated by a single time unit. For each age class there is no longer simply a net birth rate but a death rate mi and a birth rate bi for female children. The number of newborn girls at time t + 1 is N0(t + 1) = ∑ biNi(t), and the number of women in class i + 1 is Ni+1(t + 1) = (1 − mi)Ni(t), the number surviving from class i at time t. Linear algebra is a natural tool for handling this model. Demographers frequently assume that bi and mi are independent of N, because they are interested in relatively short term predictions.
Seasonality may be important for short term models, since in many species births occur during a particular season and death rates are also dependent on the time of year. An explicit time dependence must be built into r(N) to allow for seasonal effects. In a long term model encompassing many years we could probably avoid this complication by averaging birth and death rates over an entire year.
I hope this discussion makes it clear that we can’t formulate an adequate model unless we know (1) what we hope to obtain from the model and (2) how complicated a model we are willing to tolerate. The latter is practically the same as how much data we are willing to supply, since complexity and data demands usually grow simultaneously.
1.5. ANOTHER EXAMPLE
The manager of a large commercial printing company asks your advice on how many salespeople to employ. Qualitatively, more salespeople will increase sales overhead, while fewer salespeople may mean losing potential customers. Thus there should be some optimum number. By salespeople
I don’t mean clerks, but people who travel, selling a company’s products to other businesses; however, these ideas could be applied to salesclerks, too. This problem has been adapted from A. A. Brown et al. (1956). The original paper goes into greater depth than the following discussion and is well worth reading.
The problem as stated is unanswerable. What are the production limitations of the company? What are the goals of management? Maximum profit? Maximum empire
with satisfactory profit? Something else? Unless these and similar questions are very clearly answered, recommendations may be quite inaccurate. A better approach would be to provide a description of the consequences of sales forces of various sizes. This would leave the final decision up to management, which is as it should be. To determine what effect a sales force will have, we must know what salespeople accomplish. Thus we can try to determine how salespeople spend their time and what results they obtain as a consequence of spending their time in that way. As long as salespeople need to be studied, we may as well ask: What is the best way (in terms of obtaining sales) for them to spend their time? We can then advise management on (1) how to obtain the greatest return from their sales force, and (2) the impact various sizes of sales forces will have on sales. This tentatively completes step 1.
Notice that we have changed the original problem considerably. We were asked, How many salespeople should be employed?
Instead, we are going to answer two other questions which we formulated at the end of the previous paragraph. Actually the questions need further refinement. For example, different salespeople have different abilities, and their territories are probably different. The question how salespeople should spend their time contains a trap; because it invites us to ignore these variations. Again, if we change the size of the sales force, we can change the total geographical area covered, the effort expended per customer, or both. Thus the question on the consequences of various sizes of sales forces also contains potential traps. Clearly step 1 hasn’t been completed; however, the best idea is probably to proceed and to realize that in studying a real situation we will eventually need to return to step 1 and formulate the questions more precisely in a way that depends greatly on the particular printing company being studied.
The major factor that will affect how much time a salesperson spends on a customer is what the salesperson can hope to gain. Observations indicate that businesses normally place most of their printing orders with one company. Hence we can classify customers as in hand
or potential.
The former need to be held, and the latter need to be converted. In addition, we can classify customers according to how much money they have to spend. As an approximation we can assume (but it should be checked) that holding and conversion probabilities are independent of size. By running an experiment with the salespeople, or possibly by examining records if we are lucky, we can obtain an idea of how conversion and holding probabilities vary with the amount of time per week devoted to a customer. From this we can decide how a salesperson should spend their time, because one additional hour per month should produce the same expected gain in revenue regardless of which customer it is spent with. (If you don’t see this, don’t worry, I’ve omitted some details. Try rereading it after Chapter 4.) This completes steps 2 and 3 for the first part of the problem. We don’t have the data to carry out step 4, but it should