Isikhathi sokufunda esilinganiselwe: 10 imizuzu
Ukukhula kokutshalwa kwezimali ezinhlelweni ze-AI kanye nokusetshenziswa okukhulayo kwe-AI endaweni yebhizinisi kuyinkomba yokuthi imakethe yemisebenzi ithuthuka kanjani, kochwepheshe be-AI.
Ubuhlakani bokwenziwa cishe bungenye yentuthuko ejabulisa kakhulu esibhekana nayo njengabantu. Kuyigatsha lesayensi yekhompiyutha elizinikele ekudaleni imishini ehlakaniphile esebenza futhi esabela njengabantu.
Kunezinhlobo ezine eziyinhloko ze-AI. ngingu:
Lolu hlobo lwe-AI luyasebenza futhi alunawo amandla okwenza “izinkumbulo” noma ukusebenzisa “izinto ezidlule” ukwenza izinqumo. Le mishini yakhelwe ukwenza imisebenzi ethile. Isibonelo, abenzi bekhofi abahlelekayo noma imishini yokuwasha yakhelwe ukwenza imisebenzi ethile, kodwa ayinayo inkumbulo.
Lolu hlobo lwe-AI lusebenzisa ukuzizwisa okudlule kanye nedatha yamanje ukwenza isinqumo. Inkumbulo enomkhawulo isho ukuthi imishini ayikhiqizi imibono emisha. Banohlelo olwakhelwe ngaphakathi olulawula inkumbulo. Ukuhlelwa kabusha kwenziwa ukuze kwenziwe izinguquko emishinini enjalo. Izimoto ezizishayelayo ziyizibonelo zobuhlakani bokwenziwa abanenkumbulo elinganiselwe.
Le mishini ye-AI ingakwazi ukusebenzelana nabantu futhi iqonde imizwa yomuntu futhi izoba nekhono lokuqonda umuntu ngokuqonda ngokusekelwe endaweni yakhe, izici zobuso, njll. Imishini enamakhono anjalo ayikasungulwa. Luningi ucwaningo olwenziwayo ngalolu hlobo lobuhlakani bokwenziwa.
Leli ikusasa lobuhlakani bokwenziwa. Le mishini izoba ehlakaniphe kakhulu, izwela futhi iqaphele. Bayakwazi ukusabela ngendlela efanayo kumuntu, nakuba kungenzeka babe nezici zabo siqu.
Ake sihlole izindlela ezilandelayo ezichaza ukuthi singasebenzisa kanjani ubuhlakani bokwenziwa:
Yilokhuukufunda okuzenzakalelayo okunikeza i-AI ikhono lokufunda. Lokhu kwenziwa ngokusebenzisa ama-algorithms ukuze kutholwe amaphethini futhi kukhiqizwe imininingwane evela kudatha abavezwe kuyo.
Theukufunda okujulile, okuyingxenye engaphansi yokufunda komshini, inikeza ubuhlakani bokwenziwa obunekhono lokulingisa inethiwekhi ye-neural yobuchopho bomuntu. Ingenza umqondo wamaphethini, umsindo, nemithombo yokudideka kudatha yakho.
Ake sizame ukuqonda ukuthi isebenza kanjani deep learning.
Cabangela isithombe esiboniswe ngezansi:
Isithombe esingenhla sibonisa izendlalelo ezintathu eziyinhloko ze-a inethiwekhi ye-neural:
Izithombe esifuna ukuzihlukanisa zingena kusendlalelo sokufakwayo. Imicibisholo idwetshwa esithombeni iye kumaphoyinti angawodwana kusendlalelo sokufaka. Ichashazi ngalinye elimhlophe kusendlalelo esiphuzi (ungqimba lokokufaka) limelela iphikseli esithombeni. Lezi zithombe zigcwalisa izindawo ezimhlophe kusendlalelo sokufaka.
Kufanele sibe nombono ocacile mayelana nalawa mazinga amathathu njengoba silandela lesi sifundo se-AI.
Izendlalelo ezifihliwe zibophezelekile kunoma yiziphi izibalo zezibalo noma ukukhishwa kwesici kokokufaka kwethu. Esithombeni esingenhla, izendlalelo eziboniswe ngowolintshi zimelela izendlalelo ezifihliwe. Imigqa ebonakalayo phakathi kwalezi zingqimba ibizwa ngokuthi "izisindo". Ngayinye yazo ngokuvamile imelela inombolo ye-float, noma inombolo yedesimali, ephindaphindwa ivelu kusendlalelo sokufakwayo. Zonke izisindo zibalelwa kusendlalelo esifihliwe. Amaphuzu akusendlalelo esifihliwe amelela inani elisekelwe kwisamba sezisindo. Lawa manani abe esedluliselwa kusendlalelo esilandelayo esifihliwe.
Ungase uzibuze ukuthi kungani kunamazinga amaningi. Izendlalelo ezifihliwe zisebenza njengezinye izindlela ngokwezinga elithile. Izendlalelo ezifihliwe kakhulu, kuba nzima nakakhulu idatha engenayo kanye nalokho okungakhiqizwa. Ukunemba kokuphumayo okulindelekile ngokuvamile kuncike enanini lezendlalelo ezikhona kanye nokuba yinkimbinkimbi kwedatha yokufaka.
Isendlalelo sokuphumayo sisinika izithombe ezihlukene. Uma isendlalelo sengeze zonke lezi zisindo ezifakiwe, sizonquma ukuthi isithombe siyi-portrait noma i-landscape.
Isibonelo: ukubikezela izindleko zamathikithi endiza
Lesi sibikezelo sisekelwe ezintweni ezihlukahlukene, okuhlanganisa:
Ake siqale ngedatha yentengo yamathikithi yomlando ukuze siqeqeshe umshini. Uma umshini wethu usuqeqeshiwe, sabelana ngedatha entsha ezosiza ukubikezela izindleko. Ngaphambili, lapho sifunda ngezinhlobo ezine zemishini, saxoxa ngemishini enenkumbulo. Lapha sikhuluma nje ngenkumbulo nokuthi iqonda kanjani iphethini kudatha futhi iyisebenzisela ukwenza izibikezelo zamanani amasha.
Okulandelayo kulesi sifundo sibheka ukuthi i-AI isebenza kanjani nezinye izinhlelo zokusebenza ze-AI.
Ukusetshenziswa okuvamile kobuhlakani bokwenziwa esikubona namuhla ukushintsha okuzenzakalelayo kwezinto zikagesi ezisendlini.
Uma ungena egumbini elimnyama, izinzwa ezisegumbini zibona ubukhona bakho bese zikhanyisa izibani. Lesi isibonelo semishini engenayo inkumbulo. Ezinye zezinhlelo ze-AI ezithuthuke kakhulu ziyakwazi nokubikezela amaphethini okusetshenziswa futhi zivule izinto zikagesi ngaphambi kokunikeza imiyalelo ecacile.
Ezinye izinhlelo kanye izicelo zobuhlakani bokwenziwa bayakwazi ukuhlonza izwi lakho futhi benze isenzo ngokufanele. Uma uthi “vula i-TV,” izinzwa zomsindo ku-TV zithola izwi lakho bese ziyivula.
Nge I-Mini Ikhaya le-Google ungakwenza nsuku zonke.
Ingxenye yokugcina yalesi sifundo se-AI sibonisa ukusetshenziswa kwe-AI ekunakekelweni kwezempilo.
Theubuhlakani bokufakelwa inezici ezimbalwa ezinhle zokusetshenziswa, futhi lesi sigaba sokufundisa sizokusiza ukuthi uwaqonde kangcono, kuqala ngokusetshenziswa kwe-AI ekunakekelweni kwezempilo. Isitatimende esiyinkinga siwukubikezela ukuthi umuntu unaso yini isifo sikashukela noma cha. Ulwazi oluqondile lwesiguli lusetshenziswa njengokufakwayo kuleli cala. Lolu lwazi luzobandakanya:
Buka ividiyo ka-Simplilearn ethi “Artificial Intelligence Tutorial” ukuze ubone ukuthi imodeli idalwa kanjani lesi sitatimende senkinga. Imodeli isetshenziswa nge Python usebenzisa I-TensorFlow.
Izicelo ze-Artificial intelligence zivuselelwadefiukuthi izinqubo zebhizinisi zenziwa kanjani emikhakheni ehlukahlukene, njengokumaketha, ukunakekelwa kwezempilo, izinsizakalo zezezimali nokunye. Izinkampani zilokhu zihlola izindlela ezingazuza kulobu buchwepheshe. Njengoba isifiso sokuthuthukisa izinqubo zamanje sikhula, kunengqondo ukuthi ochwepheshe bathole ubuchwepheshe ku-AI.
TheI-Artificial Intelligence yezinto (AIoT) iyinhlanganisela ye-Artificial Intelligence (AI) ngaphakathi kwezixazululo ze-inthanethi yezinto (IoT). I-inthanethi Yezinto (noma I-inthanethi Yezinto) isuselwe embonweni wezinto "ezihlakaniphile" zokuphila kwansuku zonke ezixhumene zodwa (sibonga i-inthanethi) futhi ezikwazi ukushintshanisa ulwazi olunalo, oluqoqwe kanye/noma olucutshunguliwe. .
Ngenxa yalokhu kuhlanganiswa, i-Artificial Intelligence izokwazi ukuxhuma kunethiwekhi ukuze icubungule idatha futhi ishintshisane ngolwazi nezinye izinto, ithuthukise ukuphathwa nokuhlaziya amanani amakhulu wedatha. Izicelo ezikwazi ukuhlanganisa i-IoT ne-AI zizoba ne- umthelela omkhulu ezinkampanini nakubathengi. Ezinye zezibonelo eziningi? Izimoto ezizimelayo, ukunakekelwa kwezempilo okukude, izakhiwo zamahhovisi ezihlakaniphile, ukulungiswa okubikezelwayo.
Uma sikhuluma Ukucutshungulwa kolimi ngokwemvelo sibhekisela kuma-algorithms e-Artificial Intelligence (AI) akwazi ukuhlaziya nokuqonda ulimi lwemvelo, okungukuthi ulimi esilusebenzisa nsuku zonke.
I-NLP ivumela ukuxhumana phakathi komuntu nomshini futhi ibhekana nemibhalo noma ukulandelana kwamagama (amakhasi ewebhu, okuthunyelwe ezinkundleni zokuxhumana...), kodwa futhi nokuqonda ulimi olukhulunywayo kanye nemibhalo (ukuqashelwa kwezwi). Izinjongo zingahluka kusukela ekuqondeni okulula kokuqukethwe, kuya ekuhumusheni, kuye ekukhiqizweni kombhalo ngokuzimela kusukela kudatha noma imibhalo enikeziwe njengokufakwayo.
Yize izilimi zihlala zishintsha futhi zibonakala ngezisho noma izinkulumo okunzima ukuzihumusha, i-NLP ithola izindawo eziningi zohlelo lokusebenza njengezihloli zesipelingi noma izinhlelo zokuhumusha ezizenzakalelayo zemibhalo ebhaliwe, ama-chatbots nabasizi bezwi ngolimi olukhulunywayo.
Lo Speech Recognition yikhono elivumela ikhompuyutha ukuthi iqonde futhi icubungule ulimi lwabantu ngefomethi yedatha ebhaliwe noma enye. Ngenxa yokusetshenziswa kwe-Artificial Intelligence, lobu buchwepheshe manje abukwazi ukukhomba ulimi lwemvelo kuphela, kodwa futhi namanye ama-nuances anjengama-accents, izilimi zesigodi noma izilimi.
Lolu hlobo lokuqashelwa kwezwi likuvumela ukuthi wenze imisebenzi eyenziwa mathupha ngokuvamile edinga imiyalo ephindaphindwayo, ngokwesibonelo kuma-chatbots anezwi elizenzakalelayo, ukuhambisa izingcingo ezikhungweni zokuxhumana, ngezixazululo zokubizela nokubhala ngezwi, noma kuzilawuli zokusebenzelana komsebenzisi ze-PC, iselula kanye ne-on- izinhlelo zebhodi.
TheI-General Artificial Intelligence (ngesiNgisi Artificial General Intelligence, noma i-AGI) wuhlobo lwe-AI olunamandla okuqonda, ukufunda nokwenza imisebenzi eyinkimbinkimbi. ngokufanayo nabantu.
Uma kuqhathaniswa ne-Artificial Intelligence Systems ekhethekile emisebenzini ethile (i-Narrow Artificial Intelligence noma i-ASI - Narrow AI), i-AGI ikhombisa ukusebenzisana kwengqondo, ukufunda kokuhlangenwe nakho okuhlukene, ukuqonda nokuvumelana nezimo ezihlukahlukene ngaphandle kokudinga uhlelo oluthile lomsebenzi ngamunye.
Naphezu kwebanga lamanje, inhloso yokugcina ye-AGI - nakuba kuwumsebenzi onzima - ukuya kuwo phindaphinda ingqondo yomuntu kanye namakhono okuqonda ngokuseduze ngangokunokwenzeka.
BlogInnovazione.it
I-Coveware ye-Veeam izoqhubeka nokuhlinzeka ngezinsizakalo zokuphendula izigameko zokuntshontshwa kwe-inthanethi. I-Coveware izohlinzeka ngama-forensics kanye nekhono lokulungisa…
Ukulungiswa okuqagelayo kuguqula umkhakha kawoyela negesi, ngendlela emisha nesebenzayo yokuphatha izitshalo.…
I-CMA yase-UK ikhiphe isexwayiso mayelana nokuziphatha kwe-Big Tech emakethe yezobunhloli bokwenziwa. Lapho…
Isinqumo esithi "Case Green", esakhiwe yi-European Union ukuze kuthuthukiswe ukusebenza kahle kwamandla ezakhiwo, siphothule inqubo yaso yomthetho ngokuthi...
