Kedis ku xidhan hagaha · Adag Heerka

Kediska Moodooyinka Tamarta Ku Salaysan

Codso itimaalka tamarta ku salaysan, hawlaha qaybinta, iyo dhibcaha dhibcaha, oo ay ku jirto sida tamarta qaraabada ahi u go'aamiso saamiga qaybinta caadiga ah.

Wadooyinka haga ee la xidhiidhaMoodooyinka Tamarta Ku Salaysan
Su'aal 1 ee 8

Qaabka tamarta ku salaysan, sidee tamartu ula xidhiidhaa suurtogalnimada barta xogta?

Horudhac su'aal kasta8 su'aalo

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

  1. Qaabka tamarta ku salaysan, sidee tamartu ula xidhiidhaa suurtogalnimada barta xogta?
    • Tamar sare macnaheedu waa ixtimaalka sare
    • Tamar yar macnaheedu waa ixtimaalka sare (p(x) waxay la mid tahay exp(-E(x))))
    • Tamarta iyo itimaalka maaha kuwo xidhiidhsan
    • Itimaalka waxay la mid tahay tamarta si toos ah
  2. Maxaa ka dhigaya tababarka moodooyinka tamarta ku saleysan mid adag?
    • Shaqada tamarta ma noqon karto shabakad neerfaha
    • Waxay u baahan yihiin xog calaamadeysan oo keliya
    • Ma matali karaan sawirro
    • Hawsha qaybinta Z (caadiye ka sarreeya dhammaan agabyada) inta badan waa mid aan la xisaabin karin
  3. Qaabkee muunada loo isticmaalo in laga soo qaado muunado EBM?
    • Dhaqdhaqaaqa Langevin (MCMC ku salaysan gradient)
    • Baadhitaanka Beam
    • Quicksort
    • Baas hore oo keli ah
  4. Maxay hababka sida isbarbardhigga dhibcaha uga ilaalin karaan xisaabinta shaqada qaybinta Z?
    • Sababtoo ah Z had iyo jeer waxay la mid tahay 1
    • Sababtoo ah waxay iska indhatiraan shaqada tamarta
    • Sababtoo ah gradient ee log p (x) waa -gradient ee E (x), iyo Z (joogta ah ee x) hoos u dhaco marka la kala soocayo.
    • Sababtoo ah waxay ku shaqeeyaan oo kaliya xog gaar ah
  5. Midkee kuwan ka mid ah moodel tamar ku salaysan oo caadi ah?
    • Transformer-ka
    • Geedka go'aanka
    • K-macneheedu waa algorithm
    • Shabakadda Hopfield
  6. For a differentiable EBM density with Z constant in x, how does its score relate to energy?
    • The score equals the partition function itself
    • The score equals the negative input gradient of energy
    • The score ignores the learned energy entirely
    • The score is always the positive energy gradient
  7. An EBM adds the same constant c to every E(x). What happens to normalized probabilities?
    • Every probability increases by c
    • All inputs become equally probable
    • They remain unchanged after the partition function is adjusted
    • The normalizer must be set to zero
  8. For two configurations with E(a) lower than E(b), what follows under the same normalized Gibbs model?
    • Configuration a has greater probability than b
    • Configuration b has greater probability because its energy is higher
    • Both probabilities must be exactly one
    • The comparison requires different partition functions for each point