Kedis ku xidhan hagaha · Dhexdhexaad Heerka

Kediska Barashada Xoojinta Wakiilo Badan

Baro sida dhowr wakiil oo xoojin-waxbarasho ah ay wax ku bartaan abaalmarinnada la wadaago, beddelka kuwa ka soo horjeeda, aragtida qaybeed iyo caqabadaha isuduwidda.

Wadooyinka haga ee la xidhiidhaBarashada Xoojinta Wakiilka Badan
Su'aal 1 ee 8

Maxaa ka dhigaya deegaanka 'mid aan fadhiyin' marka loo eego hal wakiil oo aragtidiisa MARL?

Horudhac su'aal kasta8 su'aalo

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

  1. Maxaa ka dhigaya deegaanka 'mid aan fadhiyin' marka loo eego hal wakiil oo aragtidiisa MARL?
    • Wakiilada kale waxay isku mar baranayaan oo bedelayaan siyaasadahooda
    • Shaqada abaalgudka si aan kala sooc lahayn ayaa dib loogu habeeyaa dhacdo kasta
    • Meesha gobolku waxay u koraan iyada oo aan xidhnayn muddo ka dib
    • Dareemayaashu waxay soo celiyaan indho-indheyn buuq badan
  2. Waa maxay macnaha 'tababarka dhexe ee fulinta baahsan' (CTDE)?
    • Wakiilada waxaa la tababaray oo ku shaqeeya hal server oo dhexe
    • Naqdiye la wadaago ayaa isticmaala macluumaadka caalamiga ah inta lagu jiro tababarka, laakiin wakiiladu waxay ku dhaqmaan indha-indheynta maxalliga ah marka la dirayo
    • Mid ka mid ah wakiilka sare ayaa soo saara amarada dhammaan kuwa kale wakhtiga runtime
    • Xogta tababarku waa mid dhexe laakiin abaal-marinaha waa la baahiyaa
  3. Qaab dhismeedka xisaabeedkee ayaa guud ahaan wakiil ka ah MDP ilaa wakiilo badan?
    • Barnaamij toos ah
    • Qaabka qarsoon ee Markov
    • Ciyaarta Markov (ciyaarta Stochastic)
    • Automon dowlad-goboleed xaddidan
  4. Goobta MARL ee iskaashato, sidee baa abaalgudka sida caadiga ah loo habeeyey?
    • Wakiil kastaa wuxuu kordhiyaa abaalgudka xun ee mid kale leeyahay
    • Kaliya wakiilka ugu dhaqsaha badan ayaa helaya abaalmarin kasta
    • Abaalmarinta waxaa loo qoondeeyaa si aan kala sooc lahayn tallaabo kasta
    • Wakiiladu waxay wadaagaan abaalgud koox wadaag ah oo ay si wadajir ah u kordhiyaan
  5. Which training approach did OpenAI Five use to improve at Dota 2?
    • Self-play against copies of its own current and past policies
    • A fixed lookup table of human moves
    • A supervised-only classifier trained on image labels
    • A single policy trained only against a static scripted opponent
  6. Waa maxay sababta ay 'shaqaynta deynta' gaar ahaan ugu adag tahay iskaashiga MARL?
    • Abaalmarintu isla markiiba way timaadaa wayna fududahay in la raadiyo
    • Way adag tahay in la go'aamiyo ficilka wakiilka ee gacan ka gaystay natiijada kooxda la wadaago
    • Waxaa jira hal wakiil oo kaliya oo helaya abaal-marin
    • Wakiiladu gabi ahaanba ma ilaalin karaan wax abaal marin ah
  7. What structural constraint does QMIX place on its mixing network?
    • It must be a linear sum with fixed weights
    • It ignores all individual agent values
    • The joint action-value is monotonic in each agent’s individual value
    • It requires communication at every action step
  8. How does Value Decomposition Networks (VDN) form the team action-value?
    • By multiplying all rewards by the number of agents
    • By selecting the largest local reward only
    • By training a separate team for each agent
    • By summing the individual agents’ action-values