Maswali yanayounganishwa na mwongozo · Ngumu Kiwango

Maswali ya Mitandao ya Capsule

Hukagua vekta za kapsuli, uelekezaji wa sehemu-kwa-mzima, maelezo ya mahali, alama za makubaliano na kwa nini uelekezaji unaorudiwa unaweza kuwa wa gharama kubwa.

Njia za mwongozo zinazohusianaMitandao ya Capsule
Swali 1 ya 8

Katika mtandao wa kapsuli, UREFU wa vekta ya pato la kapsuli unawakilisha nini?

Hakiki kila swali8 maswali

Majibu na maelezo huonekana unapojibu maswali.

  1. Katika mtandao wa kapsuli, UREFU wa vekta ya pato la kapsuli unawakilisha nini?
    • Uwezekano wa kuwa huluki iko
    • Rangi ya kipengele kilichogunduliwa
    • Kiwango cha kujifunza kwa safu hiyo
    • Idadi ya neurons kwenye capsule
  2. Ni tatizo gani na CNN za kawaida ambazo mitandao ya kapuli iliundwa mahsusi kushughulikia?
    • Mafunzo ya polepole kwenye GPU
    • Kupoteza mahusiano ya anga/sehemu nzima, k.m. kukubali sehemu za uso zilizochanika
    • Kutokuwa na uwezo wa kuchakata picha za rangi
    • Vigezo vichache sana vya kujifunza
  3. Je, jina la algorithm inayoamua ni vidonge vipi vya kiwango cha chini vinavyounganishwa na vidonge vya kiwango cha juu?
    • Uenezaji nyuma kupitia wakati
    • Ukataji wa gradient
    • Uelekezaji-kwa-makubaliano unaobadilika
    • Urekebishaji wa kundi
  4. Ni nani aliyeanzisha mtandao wa kibonge na uelekezaji wa nguvu mnamo 2017?
    • Yann LeCun na wenzake
    • Ian Goodfellow na wenzake
    • Vaswani na wenzake
    • Sabour, Frosst, na Hinton
  5. Ni nini madhumuni ya kazi ya 'boga' katika mtandao wa kapsuli?
    • Kurekebisha urefu wa vekta hadi kati ya 0 na 1 huku ukiweka mwelekeo wake
    • Ili kufuta vidonge visivyotumiwa
    • Ili kubadilisha picha kuwa kijivu
    • Ili kuongeza kina cha mtandao kiotomatiki
  6. Which result did the original 2017 CapsNet paper report as evidence for routing-by-agreement?
    • Lower memory use than every CNN on ImageNet
    • Better recognition of highly overlapping digits than a convolutional net
    • No need for a training objective
    • Perfect performance on all affine transformations
  7. In the original capsule representation, what information is carried by the direction of a capsule’s output vector?
    • The probability that the entity is absent
    • The number of routing iterations
    • Instantiation or pose parameters of the represented entity
    • The loss assigned to the previous layer
  8. In routing-by-agreement, what makes a lower capsule’s prediction agree with a higher-level capsule?
    • Their capsule names begin with the same letter
    • Their vector lengths are exactly zero
    • They occur in the same image crop
    • Their predicted and output vectors have a large scalar product