JL \J \,,,f JL €,4ill 1, JL J, X .1:1,1, J, JL \ , | by H. JUSTIN DAVIDSON
»& - [£"-i <//(?>; - x*) + if(x*)\
l-H?-ig*(Pi - z*) - g(x*) - pn+i]
»+lC + 2^=1 V)Qi + Pn+1
»c* = »6 - <f(z*) 4- XXi » jW
l+ic = g(x ) - 2 ^ - i 0/ «y
.LINEAR PROGRAMMING is a complicated tool. So is a car.
Suppose you have never seen a car before and arrive at
a driving school for your first lesson on how to operate
a car. The instructor takes you into a classroom: "This,
Sir, is a piston. The piston is connected to this next part
called a connecting rod. The connecting rod fastens to
to an eccentric called a crankshaft. The crankshaft. . ."
After the first lesson, you might decide that you never
really wanted to learn to drive a car. If it's that complicated
. . .
Fortunately, most of us weren't forced to start with
the piston to learn to drive. It wasn't truly necessary. We
started with the steering wheel, the clutch and gearshift
and, especially, the gas pedal.
The same analogy applies to linear programming. If
we start with the mathematical innards, the "cee-jay" and
the "ex-eye-jay," it may take a long time to learn to use
the tool — when the training really isn't necessary. In
the following paragraphs, we will attempt to develop a
working understanding of linear programming. But we
will take the short way and stay away from the mathematics.
2 THE QUARTERLY