引導式測驗 · 中等 等級

半監督學習測驗中的一致性正則化

測驗標籤保留擾動、一致性損失、未標記輸入、教師目標、置信閘和評估。

問題 1 的 8

主要的一致性假設是什麼?

預覽每個問題8 問題

當您參加測驗時,就會出現答案和解釋。

  1. 主要的一致性假設是什麼?
    • Any change to an input must change its class
    • Unlabeled data should be assigned random labels
    • Small label-preserving perturbations should not change the model’s prediction
    • All examples should produce identical outputs
  2. 如何在基本一致性損失中使用未標記的範例?
    • Assign it a verified ground-truth label
    • Use it only to compute classification accuracy
    • Remove it after augmentation
    • Compare predictions from two perturbed views and penalize their difference
  3. 哪一個是適當時保留標籤的圖像擾動的範例?
    • A modest crop or rotation that leaves the object class unchanged
    • Replacing the object with a different class
    • Removing all pixels containing the target
    • Changing the label at random
  4. 低密度決策邊界直覺顯示什麼?
    • Place a boundary through every cluster center
    • Place decision boundaries away from dense groups of similar unlabeled examples
    • Ignore the unlabeled distribution
    • Force every cluster to have one class
  5. 停止梯度在師生一致性設定中扮演什麼角色?
    • It removes the unsupervised loss
    • It freezes every model parameter permanently
    • It prevents the target branch from receiving the matching-loss gradient
    • It makes both views identical
  6. How does FixMatch use confidence?
    • It uses every weak prediction as a hard label
    • It sets confidence to one for all classes
    • It drops the strongly augmented view
    • It keeps a pseudo-label from a weak view only when confidence passes a threshold
  7. Why are unlabeled examples useful in this method?
    • They provide additional input structure without requiring a human label for each example
    • They guarantee the class labels are correct
    • They replace the labeled validation set
    • They eliminate the need for augmentation
  8. Which evaluation would best test a consistency method’s intended behavior?
    • Check only training loss
    • Measure prediction stability across label-preserving views and task quality on held-out labeled data
    • Test on a single augmented view
    • Count unlabeled examples without measuring outputs