Kediska Barashada Ku dayashada

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

1. In imitation learning, how does an agent acquire its behavior?
   • Barashada siyaasadda adoo koobiyeynaya bandhigyada khabiirada
   • Kordhinta abaal-marin-gacan lagu naqshadeeyay iyada oo loo marayo sahaminta aan tooska ahayn
   • Cadaadiska shabakada neural si ay ugu shaqeyso aaladaha yaryar
   • Ka soo saarida sawirada kicinta qoraalka
2. Xiritaanka akhlaaqda ayaa qaabeeya barashada ku dayashada sida dhibaatada noocee ah?
   • Koox aan la ilaalin
   • Barashada la kormeerayo ee khariidadaha-ficil-celinta
   • Jiil colaadeed
   • Abaalmarinta qaabaynta
3. Dhibaato noocee ah ayaa sababa cloning-dabeecadu inay ku guul darreysato taxanaha ficilka dheer?
   • Ku-habboonaanta shaqada abaalmarinta
   • Waxyaalaha qallafsan ee ku jira cod-bixiyaha
   • Khaladka isku dhafan ee ka yimid wareejinta qaybinta
   • Illowda masiibada ah ee fasallada hore
4. Sidee ayuu DAgger algorithm wax uga qabtaa daciifnimadaas?
   • Adiga oo khabiirka ka saaraya wareegga gebi ahaanba
   • Iyadoo la kordhinayo qodobka dhimista
   • Adiga oo ku daraya qaylo aan toos ahayn dhammaan ficillada
   • Adiga oo waydiinaya khabiirka gobolada ardaygu runtii wuu soo booqdaa oo isku daraa xogtaas
5. Waa maxay sababta barashada ku dayashada badanaa looga doorbido xoojinta barashada hawlaha sida wadista baabuurta?
   • Mudaaharaadyadu way fududahay in la bixiyo marka loo eego shaqada abaal-marinta saxda ah
   • Waxay dammaanad qaadaysaa dhaqanka ugu wanaagsan
   • Marna uma baahna xog
   • Si dhakhso leh ayay u socotaa marka la eego kaliya
6. What is a realistic limitation of behavioral cloning when the agent reaches states outside the demonstrations?
   • It is guaranteed to improve beyond the expert
   • Its behavior is not guaranteed there because those states may be out of distribution
   • It can only produce random actions
   • It automatically queries an expert at every step
7. In one DAgger iteration, what does the learner do before requesting expert labels?
   • Deletes all earlier demonstrations
   • Trains only on randomly generated labels
   • Rolls out its current policy and records the states it actually visits
   • Replaces the expert policy permanently
8. After the expert labels the states visited by the current policy, what happens to that data?
   • It is used only as an evaluation metric
   • It replaces the original observation space
   • It is discarded after one gradient step
   • It is aggregated with previous labeled data for retraining