The Life Starter
Sign up
Grade 7 · unit review

Review: Evaluate a Simple Classification Model

Use eight held-out cases with six correct predictions. Explain the accuracy and classify a reusable prediction whose label is not reusable. State one limitation.

Download reviews and answer keys

Bring the unit skills together

Use these 4 new cases to check your reasoning. Choices are temporary and are not saved to an account, sent to staff or recorded as school grades. Download your current review choices before leaving or reloading.

Unit review questions
1. Supplied training target says reusable. Classify the field in the stated supervised model.
Need a hint?
  • Distinguish input information, known teaching outcome and model output before evaluating a case.
  • The future answer should not be included as a feature when it is unavailable at prediction time.
2. Held-out fictional test: 9 predictions correct out of 12. Accuracy percentage?
Need a hint?
  • Divide correct predictions by all evaluated cases when calculating this test-set accuracy.
  • Do not discard wrong predictions from the denominator.
3. Negative guess for a positive label. Positive means reusable. Classify the prediction against its known label.
Need a hint?
  • Compare the predicted category with the supplied actual category.
  • A positive prediction is not proof that the labeled case is positive.
4. Evaluate once on withheld labeled cases. Which evidence role is described?
Need a hint?
  • Use an untouched evaluation set for the stated generalization check and disclose its scope.
  • Reusing test answers to tune a model makes that set part of model development.

The download contains your current choices and any checked feedback. Keep the separate transfer-task work in your own document; downloading its instructions does not mark it complete.

Use the skills in a new task

Use eight held-out cases with six correct predictions. Explain the accuracy and classify a reusable prediction whose label is not reusable. State one limitation.

Work on paper or in your own document. Explain the choices you made.

  • Use only supplied fictional inputs and state the relevant rule.
  • Include labeled steps or evidence for each decision.
  • Name one limitation or tempting mistake.
  • Check that the result meets the task and revise a justified error.
Compare with a model response

Accuracy is 6/8=75% for these eight cases. Predicting reusable for a not-reusable label is a false positive. A small supplied test does not establish performance across every object population or future setting.

A different response can also meet the criteria. Compare the evidence and reasoning, rather than copying the wording.