Kediska Kordhinta Xogta

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

1. Waa maxay ujeedada koowaad ee xog kordhinta?
   • Si loo balaadhiyo loona kala duwo xogta tababarka si moodalku si ka sii wanaagsan u soo koobaan
   • Si loo dedejiyo fikradda ku saabsan qaabka la tababaray
   • Si loo cadaadiyo cabbirka xogta kaydka
   • Si otomaatig ah loogu calaamadiyo tusaalooyinka aan la calaamadin
2. Midkee kuwan ka mid ah farsamada kor u qaadida sawirka caadiga ah?
   • Soo degid tartiib ah
   • Si toos ah u rogrogay
   • Calaamadaynta
   • Ka saarida
3. Muxuu turjumaada danbe ka qabtaa kordhinta qoraalka?
   • Wuxuu rogaa habka ay u kala horreeyaan erayada jumlada
   • Waxay ka saartaa dhammaan ereyada joojinta qoraalka
   • Wuxuu u tarjumaa qoraalka luqad kale iyo gadaal si uu u abuuro jumlado
   • U beddelo qoraalka isku xidhka nambarada
4. Maxay tahay sababta kordhintu u noqoto 'calaamadaynta-ilaalinta'?
   • Si loo ilaaliyo cabbirka faylka xogta si joogto ah
   • Si loo sameeyo tababarku si degdeg ah u socdo
   • Sababtoo ah moodooyinka ma akhriyi karaan calaamadaha si kale
   • Sababta oo ah beddelka summada waxay bari doontaa tusaalaha ururrada khaldan
5. Maxay qabtaa farsamada Mixup?
   • Wuxuu isku daraa laba tusaale iyo calaamadahooda hal muunad tababar
   • Weji-xidho sawirka gobol leydi ah
   • Si toos ah u raadiya siyaasadda kordhinta ee ugu fiican
   • Wuxuu ku daraa qaylo aan toos ahayn miisaanka moodeelka
6. Goorma ayay xog-ururinta si gaar ah u qiimo badan tahay?
   • Marka aad hore u lahayd xog aan xadidnayn
   • Marka xogta la calaamadiyay ay yaraato oo ururinta wax badan ay adagtahay
   • Kaliya marka qaabka si buuxda loo tababaro
   • Kaliya loogu talagalay xogta xaashida xaashida
7. What does CutMix do to an image and its training label?
   • It adds a random word to a sentence without changing the label
   • It swaps only class names in the dataset
   • It replaces a patch with a patch from another image and mixes labels by the replaced area
   • It blurs every input and keeps a single label unchanged
8. How does AutoAugment select augmentation policies in its original paper?
   • It chooses transformations from a fixed human-written list without evaluation
   • It uses test labels to relabel training images
   • It selects operations from inference-time confidence only
   • It searches for policies using validation performance as a reward signal