Q-Barashada
Q-Learning waa xoojinta barashada algorithm taaso baraysa wakiilka ficilada faa'iidada ugu fiican leh iyadoo si tartiib tartiib ah u baranaysa qiimaha dhaqdhaqaaq kasta iyada oo loo marayo tijaabo iyo khalad.
Dulmar
It matters because it can find optimal behavior without ever being told the rules of its environment.
quusid qoto dheer
Q-Learning waxay bartaa shaqo loo yaqaan Q(s, a): abaal-marin-dheer ee la filayo ee ah qaadashada tallaabo 'a' ee gobolka's' ka dibna u dhaqmo si wanaagsan ka dib. Wakiilku waxa uu bilaabaa in aanu waxba garanayn, waxa uu tijaabiyaa ficilada, waxanu eegayaa abaalgudka. Tallaabo kasta ka dib waxay ku nuuxnuuxsataa qiyaasta Q-qiimaha ee abaalmarinta hadda la helay iyo qiimaha mustaqbalka la dhimay ee ugu wanaagsan ee ay ka filayso gobolka soo socda. Muhiimad ahaan, waa 'siyaasad-ka-baxsan' iyo 'qaab-la'aan': waxay baran kartaa siyaasadda ugu fiican iyada oo sahaminta si aan kala sooc lahayn, waligeedna uma baahna tusaale ah sida adduunku isu beddelayo. Marka la eego sahaminta ku filan ee lammaane kasta oo tallaabo-goboleed ah, Q-qiyamku waxay si cad ugu biiraan qiyamka ugu wanaagsan, iyo tallaabada ugu wanaagsan ee gobol kasta si fudud waa kan ugu sarreeya Q.
Aragtida Farsamada
Xuddunta waa cusbooneysiinta Bellman: Q(s,a) <- Q(s,a) + alpha[r+gamma*max_a' Q(s',a') - Q(s,a)]. Alpha waa heerka waxbarashada, gamma waa qodobka dhimista miisaanaya abaal-marinaha mustaqbalka, iyo erey-ku-duubni waa khaladka kala duwanaanshaha ku-meel-gaadhka ah. 'Max' ee ka sarreeya ficillada soo socda waa waxa ka dhigaya siyaasadda ka baxsan oo u ogolaanaya inay bartaan siyaasadda ugu habboon ee hunguriga ah xitaa marka la sahaminayo. Sahaminta sida caadiga ah waxaa lagu maamulaa epsilon- hunguriga doorashada.
Saamaynta Istiraatijiyadeed
Qiimaha iyo miisaaniyada
Go'aamada qaab-dhismeedku waxay horseedaan waxqabadka iyo kharashka hawlgalka sannadaha.
Go'aamo cad
Waxbarashada farsamada waxay ka caawisaa kooxaha inay doortaan xidhmo sax ah, ma aha oo kaliya kan ugu cusub.
Xakamaynta tayada
Doorashooyinka injineernimada ee wanaagsan waxay yareeyaan shilalka la isku halleyn karo ee wax soo saarka.
Mustaqbalka Q-Learning
Shaxda caadiga ah ee Q-Learning waxay halgantaa marka gobolladu aad u badan yihiin in lagu kaydiyo miiska. Jihada ugu weyn waxay isku daraysaa shabakadaha neerfaha, sida Deep Q-Networks (DQN), kuwaas oo ku qiyaasa Q-qiyaasta agabyada cayriin sida pixels. Cilmi-baadhistu waxay ku sii socotaa sidii loo xasilin lahaa dib-u-ciyaarista waayo-aragnimada, shabakadaha bartilmaameedka ah, iyo kala duwanaanshiyaha sida Double DQN iyo Q-Learning qaybinta taas oo yaraynaysa eexda xad-dhaafka ah waxayna ka dhigan tahay qaybin buuxda oo soo celin ah halkii ay ka ahaan lahayd hal celcelis.
Dhaqangelinta Adduunka-dhabta ah
Wakiilada ciyaarta ciyaarta Atari (DeepMind's DQN) oo si toos ah uga baranaya ciyaarta Breakout iyo Pong
Wanaajinta wakhtiga iftiinka taraafigga ee isgoysyada si loo yareeyo wadarta wakhtiga sugitaanka baabuurka
Robot-ku-socod iyada oo loo marayo shabag ama maro halkaas oo robotku ku barto dariiqa ugu gaaban ee abaal-marinta
Qiimaynta firfircoon iyo go'aamada alaabada halkaas oo wakiilku barto ficilada kordhinaya faa'iidada muddada dheer
Khatarta & Dariiqyada Ilaalada
Hagaajinta hal bartilmaameed waxay qarin kartaa daciifnimada nidaamka ballaaran.
Kaabayaasha dhaqaalaha iyo dayactirka inta badan waa la dhayalsadaa.
Nabadgelyada iyo daldaloolada u fiirsashada ayaa kori kara marka nidaamyadu noqdaan kuwo aad u adag.
Qorshe Hawleedka Dhaqangelinta
Qeex daahida, tayada, iyo bartilmaameedyada qiimaha ka hor inta aan la hirgelin.
Benchmark marka la eego culeyska dhabta ah iyo xaaladaha xogta.
La socodka qalabka khaladaadka, leexashada, iyo saamaynta isticmaalaha.
U diyaari dib-u-noqoshada iyo dariiqyada jawaab-celinta dhacdada ka hor inta aanad miisaan.
Sii wad Sahaminta
Free newsletter
Get the daily AI briefing
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the Q-Learning quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
Hagaha xiga
Jadwalka Heerka Waxbarashada
Su'aalaha soo noqnoqda
What is Q-Learning?
Q-Learning waa xoojinta barashada algorithm taaso baraysa wakiilka ficilada faa'iidada ugu fiican leh iyadoo si tartiib tartiib ah u baranaysa qiimaha dhaqdhaqaaq kasta iyada oo loo marayo tijaabo iyo khalad. Waa muhiim sababtoo ah waxay heli kartaa habdhaqan wanaagsan iyada oo aan waligeed loo sheegin sharciyada deegaankeeda.
Maxay u taagan tahay Q-qiimaha Q(s, a)?
Q(s, a) waxay qiyaasaysaa wadarta guud ee abaal-marinta mustaqbalka ee la dhimay ee laga qaadayo tallaabo dawlad-goboleedyada iyo u dhaqmida si wanaagsan ka dib, maaha oo keliya abaalmarinta degdegga ah.
Waa maxay sababta Q-Learning loogu tilmaamay 'siyaasad ka baxsan'?
Ugu badnaan falalka soo socda macnaheedu waa Q-Learning waxay barataa qiimaha siyaasadda ugu habboon ee hunguriga ah xitaa marka wakiilku uu ku sahamiyo siyaasad dhaqan oo ka duwan.
Xeerka cusboonaysiinta, waa maxay sababta dhimista gamma ay maamusho?
Gamma (inta u dhaxaysa 0 iyo 1) waxay dhimaysaa abaalmarinta mustaqbalka; qiyamka ku dhow 1 wuxuu ka dhigayaa wakiilka mid fog, qiyamka ku dhow 0 wuxuu ka dhigayaa mid gaaban.
Waa maxay farqiga ku meel gaarka ah (TD) ee ku jira Q-Learning?
Khaladka TD waa farqiga u dhexeeya qiyaasta cusub ee la beegsanayo (abaalmarinta iyo qiimaha mustaqbalka la dhimay ee ugu wanaagsan) iyo qiyaastii hore ee Q; Cusboonaysiinta ayaa yaraynaysa farqigan.
Waa maxay sababta ay Q-Learning-ka cad ula halgamayso dhibaatooyinka waaweyn sida ciyaaraha fiidyaha ee pixels?
Shaxda la eegayo waxay u baahan tahay gelitaanka labadii tallaabo-goboleed kasta, taas oo aan la hirgelin karin marka ay gobolladu tiro ahaan balaayiin yihiin, taasoo dhiirigelinaysa qiyaasaha shabakadaha neural-ka sida DQN.