Kediska Laqabsiga Domain

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

1. Dhibaato noocee ah ayaa laqabsiga domain ugu horrayn loogu talagalay in lagu xalliyo?
   • Yaraynta tirada cabbiraadaha moodeelka
   • Dadajinta isku dhufashada matrixka ee GPU-yada
   • Hoos u dhaca waxqabadka marka xogta imtixaanku ka duwan tahay xogta tababarka
   • Soo saarista sumadaha tababarka synthetic si toos ah
2. Shabakadda domain-adversarial, maxaa lakabka rogaal celiska ah ku guulaysto?
   • Waxay ku riixdaa soosaarayaasha sifada inay ka dhigto isha iyo bartilmaameedka aan la kala saari karin
   • Waxay tirtirtaa kala-soocida domainka inta lagu guda jiro fikradda
   • Waxay labanlaabtaa heerka waxbarashada ee tababarka degdegga ah
   • Waxay u beddeshaa soo-saarka dib-u-celinta calaamado kala-saarid
3. Maxaa faa'iido gaar ah ka dhigaya la-qabsiga bogga 'aan la ilaalin'?
   • Uma baahnid tababar moodel haba yaraatee
   • Waxa kaliya oo ay ku shaqeysaa xogta sawirka
   • Waxay meesha ka saaraysaa baahida loo qabo domain isha
   • Uma baahna xog calaamadeysan oo ku jirta goobta la beegsanayo
4. Xaaladdee ayaa ah tusaalaha caadiga ah ee isbeddelka domain?
   • Qaabka loo tababaray oo lagu tijaabiyay isla xogta saxda ah
   • Kala soocida burada oo lagu tababaray hal sawir oo ku guul daraystay iskaanka kale
   • Qaab si degdeg ah u orda ka dib marka la qiyaaso
   • Shabakad leh lakabyo ka badan shabakad kale
5. Waa maxay habka laqabsiga xayndaabka 'ilo-free' ee loogu talagalay in laga fogaado?
   • Helitaanka xogta tababarka isha asalka ah inta lagu jiro laqabsiga
   • Isticmaalka qaab dhismeedka shabakad kasta oo neerfaha ah
   • Soo saarista saadaal kasta oo wax soo saar ah
   • Ku shaqeeya GPU-yada casriga ah
6. A team starts adaptation from a broadly pretrained model. What should it still measure?
   • Only the parameter count
   • Only source training accuracy
   • Performance on the actual target data and task
   • Whether target evaluation can be skipped
7. Which DANN head supplies the objective to distinguish source examples from target examples?
   • The task label vocabulary
   • The domain classifier
   • The optimizer checkpoint file
   • A fixed output embedding dimension
8. Does gradient reversal mean changing a positive class label into a negative class label?
   • Yes; every task label is negated
   • Yes; all target samples are relabeled as source
   • No; it freezes every feature weight
   • No; it reverses the domain gradient into the feature extractor