Information Technology Reference
In-Depth Information
Actual
Predicted
75 years ago
50 years ago
25 years ago
Now
FIGURE 8.9
You can validate a model's ability to predict 25 years into the future by using it
to “predict the present” with data 25 or more years old. You can then compare the model's
prediction of the present with current reality.
Figure 8.9 illustrates how a model can predict the present. Suppose you want to see
how well your model predicts events 25 years into the future. You have access to data
going back 75 years. You let the model use data at least 25 years old, but you do not let
the model see any data collected in the past 25 years. The job of predicting the present,
given 25-year-old data, is presumably just as hard as the job of predicting 25 years into
the future, given present data. The advantage of predicting the present is that you can
use current data to validate the model.
A final way to validate a computer model is to see if it has credibility with experts
and decision makers. Ultimately, a model is valuable only if it is believed by those who
have the power to use its results to reached a conclusion or make a decision.
The field of software engineering grew out of a growing awareness of a “software crisis.”
In the 1960s, computer architects had taken advantage of commercial integrated circuits
to design much more powerful mainframe computers. These computers could execute
much larger programs than their predecessors. Programmers responded by designing
powerful new operating systems and applications. Unfortunately, their programming
efforts were plagued by problems. The typical new software system was delivered behind
schedule, cost more than expected, did not perform as specified, contained many bugs,
and was too hard to modify. The informal, ad hoc methods of programming that worked
fine for early software systems broke down when these systems reached a certain level of
complexity.
Software engineering
is an engineering discipline focused on the production of
software, as well as the development of tools, methodologies, and theories supporting
software production. Software engineers follow a four-step process to develop a software
