Maswali yanayounganishwa na mwongozo · Ngumu Kiwango

Maswali ya Faragha ya Tofauti

Fanya mazoezi ya kutofautisha usikivu wa faragha, utunzi, na DP-SGD, ikijumuisha kwa nini mipaka rasmi ya upotezaji wa faragha inatofautiana na ahadi za kutokukariri.

Njia za mwongozo zinazohusianaFaragha ya Tofauti
Swali 1 ya 8

For otherwise comparable differential-privacy mechanisms, what does a smaller epsilon specify?

Hakiki kila swali8 maswali

Majibu na maelezo huonekana unapojibu maswali.

  1. For otherwise comparable differential-privacy mechanisms, what does a smaller epsilon specify?
    • A larger dataset file
    • An increased model parameter count
    • A tighter bound on the allowed privacy loss
    • A faster GPU clock
  2. Why is noise calibrated to sensitivity in a differential-privacy mechanism?
    • To encrypt every raw record
    • To limit how a permitted neighboring-data change affects released output distributions
    • To remove the need to define neighboring datasets
    • To guarantee every person is impossible to identify under all conditions
  3. Ni nini kinachotofautisha mtindo wa 'ndani' wa utofautishaji wa faragha kutoka kwa mtindo wa 'kati'?
    • Mitaa inahitaji mtunza mkuu anayeaminika kikamilifu
    • Wenyeji hawatumii kelele hata kidogo
    • Karibuni hufanya kazi kwenye picha pekee
    • Local huongeza kelele kwenye kifaa cha kila mtu kabla ya data kuondoka
  4. 'Usikivu' hutumika kwa nini wakati wa kurekebisha kelele?
    • Kupima ni kiasi gani cha data ya mtu mmoja kunaweza kubadilisha matokeo ya swali
    • Kuhesabu idadi ya safu wima kwenye jedwali
    • Kugundua joto la vifaa
    • Kuchagua fonti kwa ripoti
  5. Kwa nini kila swali jipya linatumia kutoka kwa 'bajeti ya faragha'?
    • Hoja hugharimu pesa kwa kila safu
    • Upotevu wa faragha hujilimbikiza katika uchanganuzi nyingi
    • Seti ya data hupungua kwa kila hoja
    • Bajeti ni kuhusu vikomo vya kuhifadhi pekee
  6. What does correctly configured DP-SGD aim to bound through clipping, noise, and privacy accounting?
    • Every possibility of memorizing any fact
    • The total number of model layers automatically
    • The need for a training objective
    • The privacy loss associated with the specified neighboring training datasets
  7. A count query includes each person at most once. Under adding or removing one person, what is its sensitivity?
    • Moja
    • The entire number of database columns
    • Zero because it is a count
    • The model’s hidden dimension
  8. Two separately randomized analyses use the same private dataset. Why evaluate them together?
    • The second analysis deletes the first output
    • Each analysis makes the previous privacy cost exactly zero
    • Privacy loss can compose across the released outputs
    • Composition applies only to public data