App Multiple Linear Regression

ivan gabrovski

Iqukethe izikhangiso

Mayelana nalolu hlelo lokusebenza

Ukuhlehla Okunemigqa Eminingi kuyindlela yezibalo esetshenziselwa ukulingisa ubudlelwano phakathi kwenguquko eyodwa exhomekekile kanye neziguquguquko ezimbili noma ngaphezulu ezizimele ngokufaka i-equation eqondile kudatha ebonwe. Ukuhlehla okunemigqa eminingi kuchaza ukuthi izibikezeli eziningana zithinta kanjani ngasikhathi sinye inguquko yomphumela.

Izingxenye eziyinhloko zokuhlehla okunemigqa eminingi:

- Inguquko exhomekekile (Y): Lena yinguquko esifuna ukuyibikezela. Ivame ukubizwa nangokuthi "inguquko eqondiwe" noma "impendulo."

- Iziguquguquko ezizimele (X1, X2, ..., Xn): Lezi yiziguquguquko esizisebenzisayo ukubikezela inguquko encikekile. Zivame ukubizwa nangokuthi "izibikezeli" noma "iziguquguquko ezichazayo."

- Imodeli yokuhlehla: I-equation yokuhlehla okunemigqa eminingi inesimo esilandelayo:

Y = beta_0 + beta_01* X1 + beta_2*X2 + ... + beta_n* Xn
lapho:
Y iyinguquko encikekile. X1, X2, ..., Xn yiziguquguquko ezizimele.
beta_0 iyi-constant (intercept). i-beta_1, i-beta_2, ..., i-beta_n yizi-coefficients zokubuyela emuva ezibonisa ithonya leziguquguquko ezizimele ezihambisanayo ku-variable exhomekekile.

Izicelo: - Ezomnotho (ukubikezela imali engenayo);- Ukunakekelwa kwempilo (ukuhlaziywa kwezici eziyingozi); -Ubunjiniyela; - Isayensi yezenhlalo; -Ukubikezela ibhizinisi.
Isibonelo: Ukubikezela intengo yendlu ngokusekelwe ku: -Usayizi wendlu; -Inani lamakamelo okulala; -Ubudala bendlu
Kuhlelo lokusebenza into ngayinye ethi Object_k(object_1, object_2 ... object_m) ichazwa ngeziguquguquko ezizimele(Xki – izici, i = 1...n ) kanye ne-variable eyodwa encikekile(Yk-target). Indlela efana nezikwele ezijwayelekile ezincane (OLS) isetshenziswa ukubala amanani afanele kakhulu ama-coefficients (beta_0, beta_1, beta_2, ..., beta_n). Inani eliqondiwe libalwa ngo:
Y = beta_0 + beta_01* P1 + beta_2 *P2 + ... + beta_n* Pn
lapho: P1, P2...Pn ziyizibikezelo ze-target.
Uhlelo lokusebenza lugcina idatha yamamodeli amaningi okubuyela emuva ku-database (DB) uhlobo lwe-SQLite olubizwa ngokuthi i-AppMultipleLinearRegression.db. Amamodeli okubuyela emuva ahlukaniswa ngamagama.

Isikrini sokuqala sohlelo lokusebenza (i-App Multiple Linear Regression Solver) sibonisa uhlu lwamasampula amamodeli okubuyela emuva (ohlwini lwe-spinner) kanye nezinkinobho zokuvumela imisebenzi ukudala (Isampula entsha), ukulayisha (Layisha), ukulondoloza (Londoloza), ukulondoloza njenge(Londoloza njenge), ukubala (Bala), nokususa (Susa) amasampula amamodeli okubuyela emuva. Kusukela esikrinini esiyinhloko, ngezakhi zemenyu, ungafinyelela nemisebenzi efana nokukhethwa kolimi, ukulondoloza nokukopisha i-database, ukuqalisa i-database ngedatha yesampula, kanye nemisebenzi yokusiza njengosizo lohlelo lokusebenza, izilungiselelo, kanye nesixhumanisi sewebhusayithi enencazelo yazo zonke izinhlelo zokusebenza ngababhali.

Imisebenzi yokudala (Isampula entsha) ifaka ibhokisi lokufaka usayizi we-matrix lapho kufakwa khona idatha yesampula entsha - inani lemigqa (inombolo ifaka umugqa wedatha ebikezelwe P1, P2...Pn- umugqa wokugcina) kanye nenani lamakholomu (inombolo ifaka ikholomu yedatha encike ku-Y1, Y2,...Yk- ikholomu yokugcina). Bese kukhiqiza ithebula lokufaka idatha efanele. Ithebula eligcwele abantu kumele liqanjwe ngaphambi kokulondolozwa. Umsebenzi Layisha sula ithebula.

Ithebula elidala eligciniwe lingase liboniswe ngokukhetha ohlwini lwe-spinner. Ithebula eliboniswayo lingabalwa futhi isixazululo sivele engxoxweni Imiphumela yohlelo lokusebenza. Umsebenzi Phrinta ungasetshenziswa kusukela kule ngxoxo kufayela le-AppMultipleLinearRegressionSolver.txt. Ifayela le-Print lifaka umsebenzi Gcina ifayela le-Db/Save ngalo likhethiwe ifolda lapho ungagcina khona ifayela. Ngemva kokukhetha ifolda kuvela inkinobho yokugcina. Kusukela emsebenzini ofanayo kungabonakala okuqukethwe kwefayela elikhethiwe, ukuqamba kabusha ifayela noma ifolda, ukudala ifolda entsha kanye nokususa ifayela elikhethiwe.
Ukuhlehla okunemigqa eminingi kuyithuluzi lokuhlaziya idatha elinamandla, kodwa kumele lisetshenziswe ngokuqapha nokuqonda imikhawulo yalo.
Okubi: Kuzwela ku-multicollinearity (ukuxhumana okuqinile phakathi kweziguquguquko ezizimele). Akuhlali kubamba ubudlelwano obungeyona imigqa. Kudinga ukuqinisekiswa ngokucophelela nokuhlolwa kwezicabango.