Maswali yanayounganishwa na mwongozo · Kati Kiwango

Maswali ya Kueneza Nyuma

Jaribu viwango vya uenezi wa nyuma, mpangilio wa utegemezi wa kinyume, thamani zilizohifadhiwa mbele, kuzidisha kanuni za mnyororo, hesabu za njia zinazoshirikiwa na utengano wa viboreshaji.

Njia za mwongozo zinazohusianaUenezaji wa nyuma
Swali 1 ya 8

What quantity does backpropagation compute for a neural network's relevant weights?

Hakiki kila swali8 maswali

Majibu na maelezo huonekana unapojibu maswali.

  1. What quantity does backpropagation compute for a neural network's relevant weights?
    • Derivatives of the loss with respect to those weights
    • A causal percentage of total error uniquely assigned to each weight
    • The architecture drawing without any calculation
    • A permanently updated set of weights without an optimizer
  2. Ni kanuni gani ya hisabati iliyo katika moyo wa uenezaji wa nyuma?
    • Nadharia ya Pythagorean
    • Nadharia ya Bayes
    • Kanuni ya mlolongo wa calculus
    • Fomula ya quadratic
  3. Why can reverse-mode backpropagation compute many parameter derivatives efficiently?
    • It avoids all local derivatives
    • It reuses graph structure and relevant intermediate results
    • It calculates only the output prediction again
    • It guarantees every implementation has exactly one forward pass of runtime
  4. Je, uenezaji wa nyuma unakokotoa viwango vipi?
    • Kutoka kwa safu ya pembejeo kuelekea safu ya pato
    • Tabaka zote kwa wakati mmoja bila utaratibu
    • Tu kwa safu ya kati iliyofichwa
    • Kutoka kwa safu ya pato kwenda nyuma kuelekea safu ya ingizo
  5. Does computing backpropagation gradients by itself establish that model weights were updated?
    • Yes; every backward computation automatically applies a learning rate
    • Yes; gradient values always replace the weights directly
    • No; the optimizer update is a separate step
    • Only if the network has no parameters
  6. In a conventional cached-backpropagation workflow, what does the forward computation provide for later derivative calculations?
    • Relevant intermediate values that can be retained for the backward computation
    • A completed optimizer update without any update rule
    • A guarantee that all later evaluation errors disappear
    • A permanent deletion of the training objective
  7. On one composed backpropagation path, local scalar derivatives are three and four. What is their chain-rule product?
    • Saba
    • Moja
    • Three quarters
    • Kumi na mbili
  8. A shared parameter has loss-gradient contributions two and five through two paths. What combined contribution is accumulated?
    • Kumi
    • Saba
    • Tatu
    • Two fifths