Basics GUIDE
Neural Networks
A neural network imhando yekudzidza-muchina yakagadzirwa neakabatana masvomhu maitiro ane zvinogadziriswa paramita.
Teerera kugwaro iriAI-yakarondedzerwa · ~3 min
Pa peji ino3 min verenga
Pfupiso
Matanho anoshandura iyo inopinza kuita inobuda, uye kudzidziswa kunogadzirisa iwo ma paramita kuti avandudze mashandiro pachinangwa chakasarudzwa.
Zvakakosha zvinobuda
- Huremu uye zvakarerekera zvinodzidzwa parameters; activation mabasa anoshandura mhedzisiro yepakati.
- Backpropagation inoverengera magradients anoshandiswa ne optimizer.
- Iyo yemukati activation haisi yega mukana kana tsananguro.
Kudzika Kwakadzika
A basic artificial neuron inosanganisa kukosha kwekuisa uchishandisa uremu, inowedzera kurerekera, uye inoshandisa activation basa. Huremu hunodzora kuti kupinza kwega kwega kunopa zvakadii. Kurerekera kunoshandura mhedzisiro. A nonlinear activation inoita kuti layers imiririre hukama hwaisakwanisa kuita murwi wemitsetse chaiwo. Semuenzaniso, iyo ReLU activation inodzosa zero yekuisa isina kunaka uye inosiya yakanaka yekuisa isina kuchinjwa. Manetiweki anogona kushandisa akasiyana activation mumatanho akasiyana. Chikamu chinobuda chinosarudzwa kuti chienderane nebasa: kufanotaura kwenhamba hakududzirwi nenzira imwechete nezvibodzwa zvezvikamu zvinogoneka. Panguva yekudzidziswa, basa rekurasikirwa rinofananidza kuburitsa nemugumisiro unodiwa. Backpropagation inoshandisa mutemo weketani kuverenga kuti maparameter anokanganisa sei kurasikirwa. An optimizer anobva ashandisa iyo ruzivo kugadzirisa ma parameter. Backpropagation computes gradients; Haisi vimbiso yekuti modhi ichawana mhinduro yakanakisa kana kuita generalize zvakanaka. Kufananidza kwehuropi kunogumira. Artificial neurons masvomhu abstractions, uye network inobudirira haisi humbowo hwepfungwa dzakafanana nemunhu. Network hombe inogona kuenzanisira hukama hwakaoma, asi inogona zvakare kudhura yakawanda kumhanya, kukwana mapatani asina basa, kana kutadza kana mamiriro achinja. Enzanisa neyakareruka yekutanga uye bvunzo pamienzaniso kunze kweiyo data yekudzidziswa.
04Muenzaniso wakashanda
Verenga imwe artificial neuron
Shandisa mapinjisi maviri, 0.8 uye 0.5, ane huremu 0.6 uye -0.4 uye bias ye0.1.
Nhamba yakayerwa ndeye (0.8 × 0.6) + (0.5 × -0.4) + 0.1 = 0.38.
ReLU inodzoka 0.38. Kana iyo yechipiri yekupinda ikachinja kuita 1.5, iyo sum inova -0.02 uye ReLU inodzorera 0.
Zvazvinoratidza
Ichi chiverengero chekuenzanisa ishanduko imwe mukati metiweki. Kukosha 0.38 ndeye activation, kwete 38% yekuvimba.
Kuitwa muNyika yeChokwadi
Network network inoshandura mapixel values kuita maficha anobatsira pakuronga mufananidzo.
Mutauro modhi inoshandura maratidziro ezviratidzo kuita zvibodzwa zvinoshandiswa kugadzira anotevera tokens.
Manetiweki ekufungidzira anoburitsa zvakaonekwa zvichangobva kuitika kune fungidziro yenhamba iyo inofanirwa kuongororwa maererano nezvaizobuda.
Kuzvishandisa
Nei zvichikosha
- Sarudzo dzakajeka. Inokubatsira kuparadzanisa zvakajeka zvichemo zvehunyanzvi kubva mumutauro wekushambadzira.
- Mutengo uye bhajeti. Iwe unogona kubvunza zvirinani kuita mibvunzo usati washandisa mari kana nguva.
- Team uye workflow. Zvikwata zvine nzwisiso yakagovaniswa inoita zvirinani chigadzirwa, mutemo, uye sarudzo dzekudzidza.
Chenjera
- Zvikwata zvakasiyana zvinogona kushandisa izwi rimwechete zvakasiyana, saka tsanangura nzvimbo nekukurumidza.
- Benchmarks inogona kutaridzika yakasimba nepo chaiyo-yenyika kuita isina kuenzana.
- Kuregeredza mhando yedata uye zvirongwa zvekuongorora zvinowanzogadzira mhedzisiro isina kusimba.
Matanho anotevera
- Tanga netsanangudzo yemutauro wakajeka yemhedzisiro yaunoda.
- Sarudza metric imwe yekubudirira uye imwe yekutadza mamiriro usati waedzwa.
- Mhanya mutyairi mudiki ane data remumiriri, kwete demo rakakwenenzverwa.
- Nyora apo Neural Networks inobatsira uye uko nzira dzakareruka dziri nani.
Sources uye kuwedzera kuverenga
Ramba Uchiongorora
Tsamba yemahara
Tora iyo yemazuva ese AI pfupiso
Nhatu dzakasimbiswa dzeAI nyaya mangwanani ega ega pakati pevhiki, dzakanyorwa neChirungu chakajeka. Yemahara zvachose, hapana ads.
Imwe email zuva rega rega revhiki. Bvisa kunyoresa kamwe chete. Hatimbotengese kana kugovera kero yako.
Zviedze iwe pachako
Take the Neural Networks quiz
Mhinduro yepakarepo pamhinduro yega yega, uye chitupa chinogovaniswa chine ID inobatika kana wapedza kosi.
Tsigira dzidzo yemahara yeAI. AI Understanding ndeye 501(c)(3) isingabatsiri - hapana ads, hapana paywall, zvachose. Ita chipo
Mibvunzo inowanzo bvunzwa
Nei neural network ichida activation mabasa?
Nonlinear activation mabasa regai akaturikidzana akaturikidzana anomiririra nonlinear hukama. Kurongedza chete mitsara mashandiro kwaizoramba kuchigadzira mutsara shanduko.
Iyo yakakura neural network inogara iri nani?
Kwete. Kuita kunoenderana nebasa, data, kudzidziswa, kuongorora, uye zvipingamupinyi zvekutumira. Mamwe ma parameter anogona kuwedzera mutengo uye haavimbisi zvakavimbika zvinobuda.
Ramba uchidzidza
Mazano anoenderana
Mamwe madhairekitori akasarudzirwa nyaya iyi