Idanwo ti o sopọ mọ itọsọna · Lile Ipele

Kapusulu Networks adanwo

Checks capsule vectors, part-to-whole routing, pose information, agreement scores and why iterative routing can be costly.

Jẹmọ itọsọna awọn ọnaAwọn nẹtiwọki Capsule
Ibeere 1 ti 8

Ninu nẹtiwọọki capsule kan, kini PIGUN ti fekito iṣelọpọ capsule kan duro?

Awotẹlẹ gbogbo ibeere8 ibeere

Awọn idahun ati awọn alaye han bi o ṣe n mu ibeere naa.

  1. Ninu nẹtiwọọki capsule kan, kini PIGUN ti fekito iṣelọpọ capsule kan duro?
    • Awọn iṣeeṣe ti nkankan ti wa ni bayi
    • Awọn awọ ti ẹya-ara ti a ri
    • Oṣuwọn ẹkọ fun ipele yẹn
    • Nọmba awọn neuronu ninu capsule
  2. Iṣoro wo ni awọn CNN boṣewa jẹ awọn nẹtiwọọki capsule ti a ṣe ni pataki lati koju?
    • Ikẹkọ lọra lori awọn GPU
    • Pipadanu awọn ibatan aaye/apakan-gbogbo, fun apẹẹrẹ. gbigba jumbled oju awọn ẹya ara
    • Ailagbara lati ṣe ilana awọn aworan awọ
    • Awọn paramita diẹ diẹ lati kọ ẹkọ
  3. Kini orukọ algorithm ti o pinnu iru awọn capsules ipele-isalẹ ti o sopọ si iru awọn capsules ipele giga?
    • Backpropagation nipasẹ akoko
    • Idinku gige
    • Yiyipo afisona-nipasẹ-adehun
    • Ipese deede
  4. Tani o ṣe afihan nẹtiwọọki capsule pẹlu ipa-ọna agbara ni ọdun 2017?
    • Yann LeCun ati awọn ẹlẹgbẹ
    • Ian Goodfellow ati awọn ẹlẹgbẹ
    • Vaswani ati awọn ẹlẹgbẹ
    • Sabour, Frosst, ati Hinton
  5. Kini idi iṣẹ 'squash' ni nẹtiwọki capsule kan?
    • Lati ṣe deede gigun fekito si laarin 0 ati 1 lakoko ti o tọju itọsọna rẹ
    • Lati pa awọn capsules ti ko lo
    • Lati yi awọn aworan pada si iwọn grẹy
    • Lati mu ijinle nẹtiwọki pọ laifọwọyi
  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