Metrics that match the decision on the AI Learning track. Accuracy hides failure on rare classes. A fraud model can be 99 percent accurate by never catching fraud. Choose precision, recall, or a cost-weighted score that matches the decision.
This lesson assumes you already worked through Early stopping.
The idea in practice
Write the cost of a false alarm and a miss. Pick the metric that reflects that cost. Report the base rate next to it.
A concrete check
goal = {
'track': 'AI Learning',
'lesson': 'Metrics that match the decision',
}
checks = [
'input available at decision time',
'score matches the real decision',
'failure case written down',
]
print(goal['lesson'])
for item in checks:
print('-', item)
Run the sketch locally if you have Python. The printout is a reminder of the checks, not a trained model. Replace the strings with the real inputs from your own example before you treat it as a design.
What usually goes wrong
Leading with accuracy on a 1 percent positive class. When this happens, stop adding parameters or tools. Fix the check, the data, or the permission, then run the same example again.
What to write down
- The input you are allowed to use at decision time.
- The output and the score or pass rule.
- One failure you will test on purpose.
- What you will not claim the system can do.
Practice
For a rare defect detector, say whether you care more about missing a defect or about false alarms, and name the metric.
Self-check
- Say Metrics that match the decision in one sentence that mentions an input and an output.
- Name the failure mode in this lesson and the check that would catch it.
Done when: you can explain this lesson without the page open, and you have a written failure case.