Idanwo ti o sopọ mọ itọsọna · Lile Ipele

Deede Ṣiṣan adanwo

Practice invertible flow transformations, change-of-variables densities, and Jacobian corrections, including limits of likelihood-based anomaly detection.

Jẹmọ itọsọna awọn ọnaAwọn ṣiṣan deede
Ibeere 1 ti 8

Ohun-ini mathematiki wo ni gbogbo iyipada ni ṣiṣan deede ni?

Awotẹlẹ gbogbo ibeere8 ibeere

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

  1. Ohun-ini mathematiki wo ni gbogbo iyipada ni ṣiṣan deede ni?
    • O gbọdọ jẹ laileto, ti kii ṣe iyipada aworan agbaye
    • O gbọdọ jẹ invertible (bijective) ati iyatọ
    • O gbọdọ dinku iwọn ni ipele kọọkan
    • O gbọdọ jẹ tabili wiwa ọtọtọ
  2. Ilana wo ni o jẹ ki awọn ṣiṣan deede ṣe iṣiro awọn iṣeeṣe deede?
    • Awọn iṣẹ softmax
    • The Bellman idogba
    • Ofin Bayes lo si awọn logiti
    • Ilana iyipada-ti-ayipada ni lilo ipinnu Jacobian
  3. Kini idi ti awọn faaji ṣiṣan bii RealNVP ati Glow lo awọn fẹlẹfẹlẹ idapọ?
    • Lati ṣe onigun mẹta Jakobu ki ipinnu rẹ jẹ olowo poku lati ṣe iṣiro
    • Lati ṣe awoṣe ti kii ṣe iyipada
    • Lati yọ iwulo fun pinpin ipilẹ
    • Lati yi awoṣe pada si GAN kan
  4. What does a tractable normalizing flow provide that an ordinary GAN generator generally does not?
    • A guarantee of perfect anomaly detection
    • A guarantee of faster sampling than every GAN
    • An explicitly evaluable model density through change of variables
    • A non-invertible mapping with no base distribution
  5. How should a team use normalizing-flow likelihoods for anomaly detection?
    • Assume every high-likelihood input is familiar
    • Treat density as the probability mass of a single point
    • Skip checking real anomaly labels
    • Validate them as a signal against relevant in-distribution and anomalous examples
  6. Ilana igbalode wo, ti a lo ninu awọn ọna ṣiṣe bii Iduroṣinṣin Diffusion 3, dagba lati inu awọn imọran ṣiṣan lilọsiwaju?
    • Backpropagation nipasẹ akoko
    • Ibamu ṣiṣan (kikọ aaye iyara ti o gbe ariwo lọ si data)
    • Àmi
    • Iwadi ina
  7. A one-dimensional flow inverse is z = x / 2. What log-Jacobian term appears in log p(x)?
    • log(2)
    • The parameter count of the network
    • log(1/2)
    • No term because the inverse is linear
  8. Why must a normalizing flow account for stretching or shrinking volume?
    • A transformed region’s density must reflect the changed volume
    • The transformation must lose half its dimensions
    • Every output point must have probability mass one
    • The base distribution must be discarded after training