Ikibazo gihuza ibibazo · Hagati Urwego

Ibigwi byinshi bitwaje intwaro

Kugenzura ubushakashatsi, gukoreshwa, kwicuza, epsilon-umururumba, ibihembo bya UCB ibihembo, Thompson icyitegererezo hamwe nibiranga imiterere.

Inzira zijyanye no kuyoboraIbigwi byinshi bitwaje intwaro
Ikibazo 1 Bya 8

Ubucuruzi bwo hagati mubibazo byamaboko menshi yitwaje amaboko hagati:

Reba buri kibazo8 ibibazo

Ibisubizo nibisobanuro bigaragara mugihe ufata ikibazo.

  1. Ubucuruzi bwo hagati mubibazo byamaboko menshi yitwaje amaboko hagati:
    • Ubushakashatsi no gukoresha
    • Kubogama no gutandukana
    • Gusobanura no kwibuka
    • Amahugurwa no gufata umwanzuro
  2. Mu bibazo by'amabandi, 'kwicuza' bivuga:
    • Ububiko bwakoreshejwe na algorithm
    • Gutandukanya ibihembo bitandukanye no guhora ukina ukuboko kwiza
    • Umubare w'intwaro urahari
    • Igihe cyo gutoza icyitegererezo
  3. Ingamba za epsilon-umururumba zikora iki?
    • Buri gihe ufata ukuboko hamwe nuburyo bwo hejuru
    • Ntuzigere ukora ubushakashatsi
    • Ahanini ikoresha ukuboko kwiza ariko igashakisha uko bishakiye hamwe na epsilon
    • Kurura ukuboko kwose inshuro zingana ibihe byose
  4. Uburyo bwo Kwiringira Hejuru (UCB) bukurikiza ihame rya:
    • Kwiheba imbere yo gushidikanya
    • Kwirengagiza gushidikanya rwose
    • Guhitamo bisanzwe
    • Icyizere imbere yikibazo
  5. Icyitegererezo cya Thompson gihitamo ukuboko na:
    • Gutoranya uhereye kuri buri kuboko kwinyuma yinyuma no gukina icyitegererezo cyiza
    • Buri gihe gutora ukuboko gukurura byibuze
    • Gutoranya intwaro muburyo buzengurutse-robin
    • Kugabanya igihombo cyamahugurwa
  6. Nigute amabandi asanzwe atandukanye nayasanzwe?
    • Irengagiza ibihembo rwose
    • Ikoresha ibiranga ibihe byubu kugirango imenyeshe guhitamo
    • Ifite ukuboko kumwe gusa
    • Ntishobora kwiga igihe
  7. In UCB, which arm usually receives a larger exploration bonus?
    • An arm whose sample count is largest
    • An arm with no uncertainty by definition
    • An arm that has been sampled relatively few times
    • An arm with a negative reward only
  8. What does logarithmic regret growth indicate over a long run?
    • The algorithm receives zero reward after each round
    • The best arm changes at every step
    • Exploration stops before the first observation
    • The cumulative gap from always choosing the best arm grows much slower than the number of rounds