Kedis ku xidhan hagaha · Dhexdhexaad Heerka

Kediska Dib-u-celinta Logistic

Waxay hubisaa itimaalka sigmoid iyo softmax, ku habboonaanta iskutallaabta, tafsiirka log-odds, xadka go'aanka, iyo doorashooyinka habaynta.

Wadooyinka haga ee la xidhiidhaDib-u-celinta Lojistikada
Su'aal 1 ee 8

In kasta oo magaciisa, waa maxay dib-u-celinta saadka ee dhabta ah ee loo isticmaalo?

Horudhac su'aal kasta8 su'aalo

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

  1. In kasta oo magaciisa, waa maxay dib-u-celinta saadka ee dhabta ah ee loo isticmaalo?
    • Kala soocida qaybaha
    • Saadaasha tirooyinka joogtada ah
    • Ururinta xogta aan la calaamadin
    • Dhimista cabbirrada
  2. Muxuu sigmoid-ka (Lojistikada) ka qabtaa qaabka?
    • Waxay soo saartaa tirada astaamaha
    • Waxay khariidad kasta oo qiimo ah u samaynaysaa itimaalka u dhexeeya 0 iyo 1
    • Waxay meesha ka saartaa waxyaabaha ka baxsan
    • Waxay uruurisaa xogta
  3. Shaqadee luminta dib-u-celinta saadka ayaa yaraysa inta lagu jiro tababarka?
    • Khalad afar gees leh celcelis ahaan
    • Inertia
    • Log luminta (cross-entropy)
    • Luminta hinge
  4. Waa maxay sababta dib-u-celinta saadka loogu qiimeeyay fasiraada?
    • Wax xisaab ah ma isticmaasho
    • Waxay qarisaa caqligeeda gudaha
    • Had iyo jeer waxay saadaalisaa boqolkiiba 50
    • Miisaan kastaa wuxuu muujinayaa sida astaantu u beddesho jaan-gooyada natiijada
  5. Sidee buu dib-u-celinta saadka caadi ahaan u beddelaa ixtimaalka calaamad fasalka?
    • Waxay khusaysaa marin, inta badan 0.5
    • Had iyo jeer waxay doorataa fasalka koowaad
    • Celcelis ahaan deriska
    • Waxay adeegsataa xulashada random
  6. Sidee dib-u-celinta saadka ugu xidhaa shabakadaha neerfaha?
    • Iyaga kuma xidhna
    • Shabakadda neerfaha sigmoid-ka ugu dambeeya ama lakabka softmax waa asal ahaan dib-u-celinta saadka
    • Waxay beddeshaa dhammaan lakabyada qarsoon
    • Waxay ku socon kartaa oo kaliya GPU-yada
  7. What does convexity tell us about the binary logistic-regression negative log-likelihood?
    • It guarantees exactly one finite solution for every dataset
    • It makes the model immune to mislabeled examples
    • A finite local minimum is global, but convexity alone does not guarantee uniqueness
    • It means the decision boundary must be nonlinear
  8. What can complete separation imply for unregularized logistic-regression maximum likelihood?
    • The model must misclassify every training example
    • The sigmoid stops mapping scores to probabilities
    • The optimizer automatically adds regularization
    • There may be no finite maximum-likelihood estimate because coefficients can grow without bound