引導式測驗 · 中等 等級
AdaBoost 測驗
追蹤 AdaBoost 的順序學習器、加權誤差、範例重新加權、最終投票以及對吵雜困難案例的敏感度。
問題 1 的 8
AdaBoost 的弱學習器通常是如何相對於彼此進行訓練的?
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。
- AdaBoost 的弱學習器通常是如何相對於彼此進行訓練的?
- Independently under unchanged sample weights
- All at once using one common unweighted loss
- Sequentially, with later rounds adapting to earlier errors
- By retaining only the final learner and discarding prior rounds
- 學習者對範例進行錯誤分類後,其相對訓練權重通常會發生什麼變化?
- It becomes identical to the learner's ensemble vote
- It increases for a subsequent round
- It is permanently removed after its first misclassification
- It is reset to equal weight before every subsequent round
- 在經典的二元公式中,什麼樣的加權誤差可以獲得有用的正向學習者投票?
- An error greater than one-half
- Any error because the ensemble weight is always positive
- An error measured only on the final test set
- A weighted error below one-half
- 為什麼錯誤標記的異常值會為 AdaBoost 帶來問題?
- They may keep receiving increased weight
- Reweighting proves that their labels must be correct
- Each stump automatically removes examples with inconsistent labels
- Repeated rounds guarantee such examples have no influence
- 一棵分裂決策樹有什麼常見的暱稱?
- Decision forest
- Decision branch
- Decision stump
- Decision ensemble
- What combines the sequence of learners into the final AdaBoost prediction?
- Only the last learner's prediction
- A weighted vote of their outputs
- A vote with necessarily equal weights for every learner
- The prediction of the learner trained on the largest subset alone
- What should a practitioner inspect when several examples remain difficult across rounds?
- Whether high weight alone proves every case is mislabeled
- Whether all difficult cases should automatically be removed
- Whether increasing rounds guarantees that all cases become representative
- Whether they are valid, mislabeled or outside intended use
- Which statement about more boosting rounds is justified?
- More rounds do not guarantee better generalization
- Lower training error guarantees lower future error
- Repeated weighting guarantees correction of noisy labels
- A large ensemble eliminates the need for held-out evaluation