# Ifomula Yokusabalalisa Okuvamile kuzibalo
Ukusatshalaliswa okuvamile, okwaziwa nangokuthi ukusatshalaliswa kweGaussian noma i-bell curve, kungenye yemiqondo eyisisekelo kakhulu kwizibalo. Ukuba khona kwayo kuvame ukubhekwa njengesisekelo sokuhlaziywa okuhlukahlukene kwezibalo kanye namathuba. Lokhu kusatshalaliswa akugcini nje ngokusetshenziswa njalo embonweni kodwa futhi nasezinhlelweni ezahlukahlukene ezisebenzayo, njengokuphathwa kwezingozi zezimali, isayensi yezenhlalo, ezokwelapha, nokuningi.
## Incazelo Yokusabalalisa Okuvamile
Ukusatshalaliswa okuvamile kuwukusatshalaliswa kwamathuba okuqhubekayo okulingana nesilinganiso sawo. Ngamanye amazwi, igrafu yesithombe yalokhu kusatshalaliswa izokwakha ijika lensimbi elikhula kusilinganiso futhi linciphe emisileni. Lokhu kusatshalaliswa kunemingcele emibili eyinhloko: isilinganiso (μ) kanye nokuphambuka okujwayelekile (σ).
Isilinganiso sinquma indawo yesikhungo sokusabalalisa, kuyilapho ukuphambuka okujwayelekile kulinganisa ukuthi idatha isakazeke kangakanani eduze kwesilinganiso. Uma ukuphambuka okujwayelekile kukhulu, ijika lokusabalalisa libanzi futhi lifushane; uma ukuphambuka okujwayelekile kuncane, ijika liyancipha futhi liqina.
## Umsebenzi Wobuningi Bamathuba
Umsebenzi wobuningi bamathuba (pdf) wokusatshalaliswa okujwayelekile unesimo sezibalo esilandelayo:
\[ f(x | \mu, \sigma) = \frac{1}{\sigma \sqrt{2 \pi}} e^{ -\frac{(x-\mu)^2}{2\sigma^2} } \]
Lapha:
– \( x \) iyi-variable engahleliwe.
– \( \mu \) yisilinganiso sokusabalalisa.
– \( \sigma \) ukuphambuka okujwayelekile kokusatshalaliswa.
– \( e \) iyisisekelo se-logarithm yemvelo, cishe u-2.71828.
Umsebenzi ongenhla udala ijika lensimbi elihambisanayo. Ukuhlanganiswa kwalo msebenzi phakathi kwamaphuzu amabili kunikeza amathuba okuthi i-random variable iphakathi kwalawo manani amabili.
## Ukusatshalaliswa Okujwayelekile Okujwayelekile
Ukusatshalaliswa okujwayelekile okujwayelekile kuwukusatshalaliswa okujwayelekile okunesilinganiso \( \mu = 0 \) kanye nokuphambuka okujwayelekile \( \sigma = 1 \). Umsebenzi wobuningi bamathuba wokusabalalisa okujwayelekile okujwayelekile uthi:
\[ f(z) = \frac{1}{\sqrt{2 \pi}} e^{ -\frac{z^2}{2} } \]
Lapha:
– \( z \) iyi-variable engahleliwe elandela ukusatshalaliswa okuvamile okujwayelekile.
Ukusatshalaliswa okujwayelekile okujwayelekile kuvame ukusetshenziswa ngoba kusenza sikwazi ukulinganisa okunye ukusatshalaliswa okujwayelekile ngenqubo ebizwa ngokuthi “ukumisa.” Ukumisa kuhilela ukuguqula amanani \( x \) okusatshalaliswa okujwayelekile \( N(\mu, \sigma) \) abe amanani \( z \) okusatshalaliswa okujwayelekile okujwayelekile \( N(0, 1) \), kusetshenziswa ifomula elandelayo:
\[ z = \frac{x – \mu}{\sigma} \]
Le nqubo yenza kube lula ukuqhathanisa amanani avela ekusatshalalisweni okujwayelekile okuhlukile ngokuwabeka esikalini esisodwa.
## Isicelo kanye Nokufaneleka
### 1. Ithiyori Yomkhawulo Ophakathi
Ukusatshalaliswa okuvamile kubaluleke kakhulu kumongo we-Central Limit Theorem (CLT). I-CLT ithi inani elikhulu ngokwanele lezinguquko ezizimele ezingahleliwe lizosatshalaliswa cishe ngokujwayelekile, kungakhathaliseki ukuma kokusatshalaliswa kokuqala. Lokhu kusho ukuthi ukusatshalaliswa okuvamile kungasetshenziswa ukulinganisa ukusatshalaliswa kwesilinganiso sesampula, uma nje isampula inkulu ngokwanele.
### 2. Isiphetho Sezibalo
Ukusatshalaliswa okuvamile kuvumela ukusetshenziswa kwezivivinyo ze-hypothesis, njenge-z-test kanye ne-t-test. Zombili izindlela zisebenzisa i-standard normal distribution ukuze kunqunywe ukubaluleka kwezibalo kwemiphumela ebonwe. I-z-test ivame ukusetshenziswa lapho usayizi wesampula mkhulu noma ukuphambuka kwesilinganiso sabantu kwaziwa, kuyilapho i-t-test isetshenziswa lapho usayizi wesampula mncane noma ukuphambuka kwesilinganiso sabantu kungaziwa.
### 3. Ukuhlaziywa Kokuhlubuka
Ekuhlaziyweni kokuhlehla okuqondile, ukucabanga ukuthi idatha yamaphutha ivame ukusatshalaliswa kubalulekile. Lokhu kucabanga kuvumela ukubalwa kwezikhawu zokuzethemba kanye nokuhlolwa kokubaluleka kwamapharamitha emodeli yokuhlehla. Ngokufanayo, ukuthola amaphutha edatha noma izinto ezingaphandle kuvame ukwenziwa ngokuhlola ukusatshalaliswa okusele ukuze kutholakale ukuphambuka okukhulu kokujwayelekile.
### 4. Ezokwelapha kanye neBiology
Kwezokwelapha, ukusatshalaliswa okuvamile kusetshenziswa ukuchaza ukusatshalaliswa kwezimo ezahlukahlukene zebhayoloji. Isibonelo, ukuphakama, umfutho wegazi, kanye nemiphumela ethile yokuhlolwa kwelebhu kuvame ukulandela ukusatshalaliswa okuvamile. Lokhu kwenza kube lula ukunqunywa kwamanani okuphela kokuxilongwa kwezokwelapha.
### 5. Ezezimali kanye Nezomnotho
Kwezezimali, ukusatshalaliswa okuvamile kusetshenziselwa ukulingisa izimo eziningi, njengokubuyiselwa kwesitoko, amazinga enzalo, nokuningi. Nakuba empeleni, amasheya avame ukubonisa ukuthambekela okuphezulu kanye ne-kurtosis, ukucabanga kokusatshalaliswa okuvamile kusanikeza isisekelo esiqinile sokuhlaziya.
## Ukusetshenziswa kanye nokubala
### Ukusebenzisa i-Python
I-Python, enamalabhulali afana ne-NumPy kanye ne-SciPy, inikeza izindlela eziningana zokusebenza ngokusatshalaliswa okuvamile. Nasi isibonelo sendlela esingahlanganisa ngayo futhi sihlele ukusatshalaliswa okuvamile sisebenzisa lawa malabhulali:
"`python
ngenisa i-numpy njenge-np
ngenisa i-matplotlib.pyplot njenge-plt
kusuka ku-scipy.stats ukungenisa okujwayelekile
# Amapharamitha okusabalalisa ajwayelekile
mu = 0 # isilinganiso
i-sigma = 1 # ukuphambuka okujwayelekile
# Idatha yokusatshalaliswa okuvamile
x = np.linspace(-5, 5, 1000)
y = norm.pdf(x, mu, sigma)
# Isakhiwo sokusabalalisa esijwayelekile
i-plt.plot(x, y)
i-plt.xlabel('x')
i-plt.ylabel('Ubuningi')
plt.title('Ukusabalalisa Okuvamile N(0, 1)')
i-plt.show()
``
Esibonelweni esingenhla, sikhiqize idatha yokusabalalisa evamile enesilinganiso esingu-0 kanye nokuphambuka okujwayelekile okungu-1, bese sidweba umsebenzi wayo wobuningi bamathuba.
## Isiphetho
Ukusatshalaliswa okuvamile kudlala indima ebalulekile kwizibalo kanye namathuba. Ukusetshenziswa kwayo yonke indawo, kusukela ku-Central Limit Theorem kuya ekusetshenzisweni okuhlukahlukene okusebenzayo njengokuhlaziywa kokuhlehla kanye nokuhlolwa kwe-hypothesis, kuyenza ibe ngenye yezindawo ezithandwa kakhulu nezibalulekile zokusabalalisa amathuba. Ukuqonda ifomula yokusabalalisa okuvamile kanye nendlela yokuyisebenzisa ngempumelelo kuyikhono elibalulekile kunoma ubani osebenza kwisayensi yedatha, ucwaningo, ezomnotho, kanye neminye imikhakha eminingi.
Ngalolu lwazi, singasondela futhi sixazulule izinhlobo ezahlukene zezinkinga zokuhlaziya ngempumelelo, okusenza sikwazi ukwenza izinqumo ezingcono ngokusekelwe kudatha etholakalayo kanye namathuba.
