Kuenderera Kudzidza uye Njodzi Kukanganwa Quiz

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

1. Chii chinonzi 'catastrophic forgetting'?
   • Kana kudzidza basa idzva kunokanganisa zvakanyanya kuita pamabasa akadzidzwa kare
   • Kana modhi inopera mundangariro uye kubondera
   • Kana data rekudzidzisa radzimwa netsaona
   • Kana modhi ikaramba kufanotaura
2. Kutengeserana pakati pemwoyo wekuenderera mberi nekudzidza kunowanzonzi:
   • Bias-variance tradeoff
   • Kugadzikana-plastiki dambudziko
   • Speed-accuracy tradeoff
   • Kuongorora-kushandiswa tradeoff
3. Chii chinoitwa neElastic Weight Consolidation (EWC)?
   • Inodzima neuroni dzisina kukosha
   • Inopeta kaviri mwero wekudzidza wemabasa matsva
   • Inoranga shanduko kune huremu hwakakosha pamabasa akadzidzwa kare
   • Inochengeta muenzaniso wese wekare wekudzidziswa zvachose
4. Nzira dzekudzokorora dzinorwisa sei kukanganwa?
   • Nekuomesa network yese zvachose
   • Nokushandisa mwero mukuru wekudzidza
   • Nokubvisa basa rekurasikirwa
   • Nekupindirana akachengetwa kana kugadzirwa ekare mienzaniso mukudzidziswa kutsva
5. How does Gradient Episodic Memory (GEM) constrain a gradient update while learning a new task?
   • It projects the gradient so its first-order approximation does not increase loss on stored prior-task examples
   • It replaces all prior-task data with new examples
   • It freezes every parameter for all future tasks
   • It increases weights that were unimportant to earlier tasks
6. Sei kusanyatsogadzirisa basa idzva kuchikonzera kukanganwa?
   • Iyo GPU inopisa uye inoshatisa ndangariro
   • Gradient descent inofambisa maremu akagovaniswa akananga kubasa idzva pasina chinomanikidza kuchengetedza ekare
   • Mabasa matsva anogara aine data rakawanda
   • Iyo modhi inodzima nemaune huremu hwekare
7. How does iCaRL classify a test example after learning multiple classes?
   • By retrieving the single closest word in a text corpus
   • By selecting the class with the largest Fisher information
   • By comparing its feature vector with per-class means computed from stored exemplars
   • By replaying every training image at inference
8. Why is evaluating continual learning more than measuring final accuracy on the newest task?
   • New-task accuracy is the only relevant measure
   • Old tasks are removed from the evaluation
   • The model cannot be tested on multiple tasks
   • A method must also measure retention on prior tasks and its ability to learn new ones