Question: Using the classifier example from the source, explain the distinction between how accuracy is treated as the optimizing metric versus how running time is treated as the satisficing metric.

Sample answer: In the classifier example, running time only needs to meet a set criterion, so it functions as a satisficing metric with a pass/fail threshold. Accuracy, by contrast, has no fixed threshold to merely satisfy; instead, the goal is to push it as high as possible. Because the source instructs maximizing accuracy subject to meeting the running time criteria, accuracy is the single metric selected for continuous improvement, making it the optimizing metric.

Key points:

• Running time has a threshold ('criteria') it must meet
• Accuracy has no threshold; it is maximized
• The source states accuracy is maximized 'subject to' the running time criterion
• This maximize-subject-to relationship defines the optimizing/satisficing distinction

Rubric: Full credit identifies that accuracy is maximized without a fixed threshold, running time only needs to meet a criterion, and correctly cites the source's phrase 'maximize accuracy, subject to... meeting the running time criteria.' Partial credit for identifying the roles without explaining the maximize-versus-threshold distinction.