Asọtẹlẹ Asọtẹlẹ

Awọn idahun ati awọn alaye han bi o ṣe n mu ibeere naa.

1. Kini ọna asọtẹlẹ ibaamu jade dipo aami ẹyọkan?
   • Eto asọtẹlẹ tabi aarin pẹlu iṣeduro agbegbe
   • Awoṣe atunṣe pẹlu iṣedede ti o ga julọ
   • A nikan ojuami ti siro pẹlu ko si aidaniloju
   • A pipadanu iye fun ikẹkọ ṣeto
2. Iru arosinu wo ni o ṣe atilẹyin iṣeduro iṣeduro ibamu deede?
   • Awoṣe jẹ nẹtiwọọki nkankikan
   • Awọn aaye data jẹ paṣipaarọ
   • Awọn ẹya ara ẹrọ ti wa ni ominira ti kọọkan miiran
   • Awọn aami jẹ iwọntunwọnsi pipe
3. Kini ipa ti eto isọdọtun?
   • Lati ṣe ikẹkọ awoṣe ti o wa labẹ
   • Lati yan oṣuwọn ẹkọ
   • Lati ṣe iṣiro awọn ikun ti ko ni ibamu ti o ṣeto iloro
   • Lati yọ awọn ita kuro lati data ikẹkọ
4. Ti o ba ṣeto alpha = 0.1, agbegbe wo ni o n fojusi?
   • 10%
   • 50%
   • 99%
   • 90%
5. Kini igbagbogbo n ṣẹlẹ si eto asọtẹlẹ nigbati awoṣe ko ni idaniloju pupọ nipa titẹ sii?
   • Eto naa gbooro sii, ti o ni awọn aami oludije diẹ sii ninu
   • Eto naa di ofo
   • Ibalona ti wa ni bikita
   • Awoṣe ti wa ni atunṣe laifọwọyi
6. Kini idi ti asọtẹlẹ ibamu ti a pe ni 'ọfẹ pinpin'?
   • O ṣiṣẹ nikan lori data Gaussian
   • Ko nilo arosinu parametric nipa pinpin data fun iṣeduro agbegbe rẹ
   • O da pinpin data silẹ patapata
   • O dawọle a aṣọ pinpin
7. A split-conformal classifier has 99 calibration examples and targets 90% marginal coverage. Which rank is used for the calibration quantile?
   • The 10th ordered score
   • The 99th score only
   • The 90th ordered calibration score (ceil((99+1) × 0.9))
   • The average of all labels
8. In classification, what can a nonconformity score such as 1 minus the candidate label probability represent?
   • Awọn nọmba ti ikẹkọ epochs
   • The probability that the calibration set is exchangeable
   • The class frequency in the test set
   • How poorly the model supports that candidate label for the example