Ikibazo gihuza ibibazo · Hagati Urwego

Ikibazo-Igihe cyo Kwongera Ikibazo

Kugenzura uburyo ikizamini-igihe cyo kongera igiteranyo gihamye-cyerekana icyitegererezo, iyo gifasha, igiciro cyacyo cyo kubara hamwe nibisabwa guhinduka.

Inzira zijyanye no kuyoboraIgihe cyo Kugerageza
Ikibazo 1 Bya 8

Mugihe cyo gufata umwanzuro, ikizamini-igihe cyo kongera gukora niki cyinjiza kimwe?

Reba buri kibazo8 ibibazo

Ibisubizo nibisobanuro bigaragara mugihe ufata ikibazo.

  1. Mugihe cyo gufata umwanzuro, ikizamini-igihe cyo kongera gukora niki cyinjiza kimwe?
    • Gukoresha icyitegererezo gihamye kuri kopi nyinshi zahinduwe zinjiza no guhuza ibisubizo
    • Kongera imyitozo kuri data yongerewe igihe cyose ibajijwe
    • Ongeraho urusaku muburemere bw'icyitegererezo mbere ya buri guhanura
    • Gucomeka icyitegererezo kugirango gikore byihuse
  2. When can pooling predictions from transformed views reduce prediction variance?
    • When every view changes the true class
    • When the transforms preserve the label and the view errors are at least partly uncorrelated
    • When all predictions are identical by design
    • When the model is retrained on each transformed copy
  3. Ni ikihe giciro nyamukuru cyo gukoresha TTA?
    • Irasaba amakuru yikizamini
    • Ihindura burundu uburemere bwatojwe
    • Yongera compte compte kuko moderi ikora inshuro nyinshi kurugero
    • Igabanya urugero rwicyitegererezo ku nyungu zisukuye
  4. Ni ubuhe buryo bwo kwiyongera bushobora gukomeretsa imibare?
    • Umucyo muto urahinduka
    • Ibihingwa bito bidasanzwe
    • Ongeraho urusaku ruto rwa Gaussiya
    • Uhagaritse gutambutsa '6' kugirango bishoboke nka '9'
  5. Mugice, niyihe ntambwe yinyongera isabwa mbere yo kugereranya ibisubizo bya TTA?
    • Guhindura buri geometrike ihinduka kugirango mask zahanuwe zisubire hamwe na pigiseli yumwimerere
    • Kongera imyitozo ya decoder kuri buri kureba
    • Kureka ibintu byose byahinduwe
    • Guhindura masike kumasanduku
  6. For a classification model, what may be aggregated across test-time views?
    • The model’s training labels
    • The predicted probabilities or logits from the views
    • The input image file names
    • The optimizer momentum buffers
  7. Does standard test-time augmentation require retraining the model on new labels?
    • Yes; each view needs a newly trained model
    • Yes; it requires extra class labels for every view
    • No; it reuses fixed weights and evaluates transformed inputs at inference
    • No; it replaces the model with a lookup table
  8. A segmentation model predicts masks on flipped and cropped views. What must happen before pixelwise pooling?
    • Average masks before undoing the transforms
    • Discard the original image coordinates
    • Apply a different label to each pixel view
    • Invert the geometric transforms so all masks align with the original image coordinates