Ikibazo gihuza ibibazo · Biroroshye Urwego

Ikibazo cyo Kongera amakuru

Ongera usuzume uburyo bwo kongera amakuru kumashusho ninyandiko, harimo ibirango-bibika impinduka, kuvanga na politiki yize.

Inzira zijyanye no kuyoboraKongera amakuru
Ikibazo 1 Bya 8

Niyihe ntego yibanze yo kongera amakuru?

Reba buri kibazo8 ibibazo

Ibisubizo nibisobanuro bigaragara mugihe ufata ikibazo.

  1. Niyihe ntego yibanze yo kongera amakuru?
    • Kwagura no gutandukanya amakuru yimyitozo kuburyo moderi rusange muri rusange
    • Kwihutisha umwanzuro kuri moderi yatojwe
    • Kugabanya ingano ya dataset kuri disiki
    • Kuranga ingero zidashyizweho ikimenyetso mu buryo bwikora
  2. Ninde muribi nubuhanga busanzwe bwo kongera amashusho?
    • Kumanuka
    • Guhinduranya bisanzwe
    • Tokenisation
    • Kureka
  3. Niki-gusubira inyuma-gukora iki mukwongera inyandiko?
    • Hindura gahunda yamagambo mu nteruro
    • Kuraho amagambo yose yo guhagarika mumyandiko
    • Hindura inyandiko kurundi rurimi hanyuma usubire gukora imvugo
    • Hindura inyandiko muburyo bwo kubara
  4. Kuki kwiyongera bigomba 'kuranga-kubika'?
    • Kugumana ingano ya dosiye yububiko
    • Kugira ngo imyitozo ikore vuba
    • Kuberako moderi idashobora gusoma ibirango ukundi
    • Kuberako guhindura ikirango byigisha icyitegererezo amashyirahamwe atariyo
  5. Tekinike ya mixup ikora iki?
    • Kuvanga ingero ebyiri na labels zabo murugero rumwe rwamahugurwa
    • Masike agace k'urukiramende rw'ishusho
    • Gushakisha politiki nziza yo kongera mu buryo bwikora
    • Ongeraho urusaku rudasanzwe kuburemere bw'icyitegererezo
  6. Ni ryari kongera amakuru bifite agaciro cyane?
    • Mugihe usanzwe ufite amakuru atagira imipaka
    • Iyo ibirango byanditse ari bike kandi gukusanya byinshi biragoye
    • Gusa iyo icyitegererezo cyatojwe byuzuye
    • Gusa kubisobanuro byurupapuro rwamakuru
  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