Imiyoboro ya Neural Isubiramo
Imiyoboro ya Neural Isubiramo (RNNs) yubatswe kugirango ikemure urutonde nkinyandiko, imvugo, hamwe nigihe cyurukurikirane.
Incamake
They process data one step at a time while carrying a memory of what came before, making order and context matter.
Kwibira cyane
Bitandukanye numuyoboro usanzwe ubona inyongeramusaruro icyarimwe, RNN isoma urukurikirane intambwe ku yindi, igaburira ibisubizo byayo kuva intambwe ibanza gusubira muri yo. Uyu muzingo ukora leta yihishe, incamake yerekana ibintu byose bigaragara kugeza ubu, ijambo "banki" rishobora gusobanurwa ukundi nyuma y "uruzi" kuruta nyuma yo "kuzigama." Ikibaya cya RNNs kirwana nurutonde rurerure kuko gradients zigabanuka cyangwa zigaturika mugihe cyamahugurwa, bigatuma bibagirwa imiterere ya kure. Amarembo ya Gatesi yakosoye ibi: Ububiko Bwigihe gito (LSTM, 1997) hamwe na Gated Recurrent Unit (GRU) byoroshye gukoresha amarembo ahitamo icyo kubika, kuvugurura, cyangwa guta, kureka umuyoboro ukagumana amakuru murwego rwinshi. RNNs yakoresheje imashini isobanura kare, kumenyekanisha imvugo, hamwe ninyandiko zahanuye mbere yuko abahindura ahanini babasimbuza.
Ubushishozi
Ibisobanuro biranga ni ibitekerezo bisubirwamo: kuri buri gihe intambwe umuyoboro uhuza ibyinjijwe hamwe na leta yabanjirije guhisha kugirango bitange leta nshya yihishe. Amahugurwa akoresha backpropagation binyuze mugihe, ikingura uruziga intambwe zose kandi ikwirakwiza amakosa inyuma. Aha niho ikibazo cyo kubura-buhoro buhoro ikibazo kiruma, kubera ko gradients yagwiriye intambwe nyinshi zerekeza kuri zeru. LSTMs yongeramo selile itandukanye hamwe ninjiza, kwibagirwa, no gusohora amarembo kugirango amakuru ashobore gutambuka mumwanya muremure hafi idahindutse.
Ingaruka z'Ingamba
Ibyemezo bisobanutse
Iragufasha gutandukanya ibyifuzo bya tekiniki bisobanutse nururimi rwo kwamamaza.
Igiciro na bije
Urashobora kubaza ibibazo byiza byo gushyira mubikorwa mbere yo gukoresha amafaranga cyangwa igihe.
Itsinda hamwe nakazi
Amakipe asangiye ibitekerezo akora ibicuruzwa byiza, politiki, nibyemezo byo kwiga.
Ejo hazaza h'urusobe rusubirwamo
Abahindura barenze RNNs kubikorwa byinshi byururimi runini kuko bitunganya bikurikiranye kandi bigafata intera ndende neza. Nyamara RNNs ntiri kure cyane: intambwe-ku-ntambwe, guhora-kwibuka-gutunganya imyenda ikwirakwiza amajwi, ibikoresho bidafite imbaraga, hamwe no kugenzura igihe. Imiterere mishya ya leta-umwanya nka Mamba kubyutsa ibitekerezo-byisubiramo-ibitekerezo hamwe nibikorwa bigezweho, bikora ibintu birebire cyane bihendutse. Tegereza inshuro nyinshi hamwe na leta-umwanya wegerejwe kugirango ugumane niche ikomeye aho amakuru ageze ubudahwema cyangwa kubara no kwibuka birakomeye.
Gushyira mu bikorwa Isi
Imbaraga hakiri kare Google Guhindura no kuvuga-sisitemu yo kwandika
Guteganya ijambo rikurikira muri terefone ya terefone ya autocomplete no kwandika swipe
Guteganya ibiciro byimigabane, ingufu zikenewe, nikirere uhereye kumateka yigihe cyuruhererekane
Kubyara no gusesengura umuziki cyangwa gutahura ibintu bidasanzwe mugukurikirana amakuru ya sensor
Ingaruka & Kurinda
Amakipe atandukanye arashobora gukoresha ijambo rimwe muburyo butandukanye, sobanura intera hakiri kare.
Ibipimo birashobora kugaragara bikomeye mugihe imikorere-yisi-itaringaniye.
Kwirengagiza ubuziranenge bwamakuru na gahunda yo gusuzuma akenshi bitanga ibisubizo byoroshye.
Igishushanyo mbonera
Tangira nururimi rusobanutse rwibisubizo ukeneye.
Toranya intsinzi imwe hamwe nuburyo bumwe bwo gutsindwa mbere yo kwipimisha.
Koresha umuderevu muto hamwe namakuru ahagarariye, ntabwo ari demo yashizweho.
Inyandiko aho Imiyoboro Yisubiramo Ifasha kandi aho uburyo bworoshye aribwo bwiza.
Komeza Ubushakashatsi
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 Recurrent Neural Networks 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
Ubuyobozi bukurikira
Shushanya imiyoboro ya Neural
Ibibazo bikunze kubazwa
What is Recurrent Neural Networks?
Imiyoboro ya Neural Isubiramo (RNNs) yubatswe kugirango ikemure urutonde nkinyandiko, imvugo, hamwe nigihe cyurukurikirane. Batunganya amakuru intambwe imwe icyarimwe mugihe batwaye kwibuka ibyaje mbere, bakora gahunda nibisobanuro.
Niki gitandukanya RNN itandukanye numuyoboro usanzwe wo kugaburira?
RNN ifite ibitekerezo bisubirwamo: buri ntambwe yihishe leta ihita imbere, igaha umuyoboro kwibuka ibice byambere byurukurikirane.
Niki 'leta yihishe' muri RNN?
Leta ihishe ikora nkurwibutso rwurusobe, ruvugururwa kuri buri ntambwe kugirango tuvuge muri make urukurikirane rwakozwe kugeza kuri iyo ngingo.
Ni ikihe kibazo gituma RNNs yibagirwa amakuru kuva kera murukurikirane?
Iyo gradients igwijwe mugihe kinini cyintambwe bakunda kugabanuka kuri zeru (cyangwa guturika), amakuru rero kare ahagarika guhindura imyigire.
Nigute LSTMs na GRUs bitezimbere kuri RNN isanzwe?
Ibice byamazu nka LSTM na GRU biga kugenzura amakuru atemba, kureka imiterere yingirakamaro igakomeza kumurongo muremure no kugabanya ikibazo cyo kubura-buhoro buhoro.
Ni ubuhe bwoko bw'amakuru RNNs yagenewe cyane cyane?
RNNs irabagirana kumibare yatanzwe aho imiterere n'ibikurikiranye, nk'interuro, imirongo y'amajwi, n'ibipimo mugihe.