Question: In a brief analytical response, explain how a target-setting framework can treat missed detections and false alarms differently for a voice-activated assistant.

Sample answer: The two error rates both matter, but they play different roles in the goal. The false alarm rate is the satisficing condition: the detector is only acceptable if it triggers no more than two false activations per day. Once a detector meets that limit, the missed-detection rate becomes the optimizing target, so the best model is the one that catches the most real wake words.

Key points:

• False alarms are activations when the wake word was not spoken.
• Missed detections are failures to activate when the wake word was spoken.
• Missed-detection rate is the quantity to minimize.
• False alarms must stay at or below two per day.
• The best detector is chosen from models that satisfy the false-alarm limit.

Rubric: A strong response correctly defines both error types, identifies missed-detection rate as the optimizing metric, states the false-alarm constraint, and explains that only models meeting the constraint are compared on the optimizing metric.