Sekela Imishini Ye-Vector
Umshini we-vector yokusekela (i-SVM) iyi-algorithm yakudala ehlukanisa amaqembu amabili ngokudweba umngcele obanzi kakhulu phakathi kwawo.
Uhlolojikelele
It was one of the most powerful classifiers before deep learning and is still strong on small, clean datasets.
I-Deep Dive
I-SVM ithola umngcele wesinqumo, obizwa ngokuthi i-hyperplane, ekhulisa umkhawulo, igebe phakathi komngcele kanye namaphoyinti edatha aseduze ekilasi ngalinye. Lawo maphuzu aseduze kakhulu 'ama-vectors asekelayo,' futhi yiwo wodwa achaza umngcele, okwenza imodeli ihlangane futhi imelane nabangaphandle kude nonqenqema. Uma idatha ingakwazi ukuhlukaniswa ngomugqa oqondile, i-kernel trick iyibeka endaweni ene-dimensional ephakeme lapho kukhona ukuhlukaniswa okuhlanzekile, ngaphandle kokuhlanganisa lezo zixhumanisi ngokuqondile. Imajini ethambile ivumela ukuhlukaniswa okuthile okungalungile, okulawulwa yipharamitha C, ngakho imodeli ibhalansisa imajini ebanzi ngokumelene namaphutha okuqeqesha. Ama-SVM ahamba kahle uma izici ziziningi kodwa izibonelo zimbalwa, njengokuhlukaniswa kombhalo kanye ne-bioinformatics.
I-Technical Insight
Ukukhulisa umkhawulo kuyinkinga yokwenza kahle i-convex, ngakho-ke ama-SVM anomsebenzi owodwa ongcono kakhulu womhlaba wonke, ngokungafani namanethiwekhi e-neural. I-kernel trick ingena esikhundleni semikhiqizo yamachashazi phakathi kwamaphoyinti edatha ngomsebenzi we-kernel, njengomsebenzi we-radial basis (RBF) noma i-polynomial kernel, ehlanganisa ukufana endaweni enobukhulu obuphakeme ngokungagunci. Lokhu kuvumela indlela yomugqa ukuthi idwebe imingcele egobile ngemali ephansi. Amapharamitha amabili abusa ukushuna: C, ehweba ngobubanzi bemajini ngokumelene namaphutha, kanye ne-gamma ku-RBF kernel, ebeka ukuthi ithonya lephoyinti ngalinye lifinyelela kude kangakanani.
I-Strategic Impact
Izinqumo ezicacile
Kukusiza ukuthi uhlukanise izimangalo ezicacile zobuchwepheshe kusukela olimini lokumaketha.
Izindleko kanye nesabelomali
Ungabuza imibuzo yokusebenzisa kangcono ngaphambi kokusebenzisa imali noma isikhathi.
Ithimba kanye nokusebenza komsebenzi
Amaqembu anokuqonda okwabiwe enza izinqumo ezingcono zomkhiqizo, inqubomgomo, nokufunda.
Ikusasa Lemishini Ye-Vector Yokusekela
Ama-SVM athathwe kakhulu ukufunda okujulile kanye nezihlahla ezithuthukisiwe ze-gradient kumadathasethi amakhulu, ayinkimbinkimbi, kodwa ahlala eyinketho ethembekile uma idatha iyivelakancane, inobukhulu obuphezulu, noma idinga isisekelo esiqinile, esiqondwa kahle. Zihlala zivamile ekufundiseni, ku-bioinformatics kanye nemisebenzi yombhalo, nasezilungiselelweni ezinomkhawulo wezinsiza lapho imodeli encane, esheshayo ishaya inethiwekhi esindayo. Lindela ama-SVM ukuthi aqhubeke njengethuluzi lakudala elithembekile kanye nebhentshimakhi kunomngcele wocwaningo olusha.
Ukuqaliswa Komhlaba Wangempela
Ukuhlelwa kombhalo nogaxekile, lapho amadokhumenti anezinkulungwane zezici zamagama kodwa izibonelo ezilinganiselwe.
Ukuhlukaniswa kwezithombe kumasethi wedatha amancane ngaphambi kokuba ukufunda okujulile kube okubusayo.
Ukuhlukaniswa komdlavuza kanye ne-gene-expression ku-bioinformatics enezici eziningi namasampuli ambalwa.
Ukuqashelwa kwedijithi ebhalwe ngesandla, ibhentshimakhi ye-SVM yakudala kudathasethi ye-MNIST.
Izingozi & Guardrails
Amaqembu ahlukene angasebenzisa igama elifanayo ngokuhlukile, ngakho chaza ububanzi kusenesikhathi.
Amabhentshimakhi angabukeka eqinile kuyilapho ukusebenza komhlaba wangempela kungalingani.
Ukuziba ikhwalithi yedatha nezinhlelo zokuhlaziya kuvame ukudala imiphumela entekenteke.
Ukuqalisa Umhlahlandlela
Qala ngencazelo yolimi olulula yomphumela oyidingayo.
Khetha imethrikhi eyodwa yempumelelo nesimo esisodwa sokuhluleka ngaphambi kokuhlolwa.
Qalisa umshayeli omncane onedatha emele, hhayi isethi yedemo ephucuziwe.
Bhala lapho i-Support Vector Machines isiza khona nalapho izindlela ezilula zingcono khona.
Qhubeka Uhlole
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 Support Vector Machines 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
Umhlahlandlela olandelayo
Izisekelo Zokufunda Ngomshini
Imibuzo evame ukubuzwa
What is Support Vector Machines?
Umshini we-vector yokusekela (i-SVM) iyi-algorithm yakudala ehlukanisa amaqembu amabili ngokudweba umngcele obanzi kakhulu phakathi kwawo. Bekungesinye sezigaba ezinamandla kakhulu ngaphambi kokufunda okujulile futhi sisaqinile kumadathasethi amancane, ahlanzekile.
Yini umshini we-vector yokusekela ozama ukuyikhulisa?
I-SVM ithola i-hyperplane eyenza imajini ibe nkulu, ibanga eliya kumaphoyinti aseduze ekilasi ngalinye, ngokuhlukana okuqine kakhulu.
Ayini 'ama-vector asekelayo' ku-SVM?
Amaphuzu aseduze nomngcele kuphela, ama-vectors asekelayo, anquma i-hyperplane; amanye amaphuzu anganyakaza ngaphandle kokuwushintsha.
Iyiphi inkinga i-kernel trick eyixazululayo?
Ubuqili be-kernel bubeka idatha ngokungaguquki endaweni enobukhulu obuphakeme lapho kusebenza khona umngcele oqondile, ovumela ukuhlukaniswa okugobile ngokushibhile.
Ilawula ini ipharamitha C ku-SVM ye-soft-margin?
Amabhalansi C anesilinganiso esikhulu ngokumelene nokuvumela amaphutha athile okuqeqeshwa; encane C isho umkhawulo obanzi, obekezelayo.
Iyiphi i-kernel evame ukusetshenziselwa ukwakha imingcele eguquguqukayo, egobile?
I-RBF (Gaussian) kernel iwukuzenzakalelayo okudumile okulinganisa ukufana ngokusekelwe ebangeni, okuvumela imingcele ebushelelezi engaqondile.