離散概率分佈 (discrete probability distribution):指所描述嘅變數
X
{\displaystyle X}
嘅可能數值係離散 嘅概率分佈[2] 。
概率質量函數 (probability mass function,PMF):描述一個離散概率分佈嘅函數;一個離散概率分佈嘅 PMF 會講明嗰個概率分佈嘅每一個離散可能數值出現嘅機會率[2] :
∑
p
X
(
x
i
)
=
1
{\displaystyle \sum p_{X}(x_{i})=1}
,啲可能性嘅機率冚唪唥加埋係 1;
p
(
x
i
)
>
0
{\displaystyle p(x_{i})>0}
,每個可能性嘅機率大過 0;
p
(
x
)
=
0
for all other x
{\displaystyle p(x)=0{\text{ for all other x}}}
,啲可能性以外嘅數值出現嘅機會率係 0。
一個概率質量函數;
X
{\displaystyle X}
嘅可能數值得三個(1、3 同 7),每個數值都掕住咗個「出現嘅機率」,而呢啲機率加埋係 1。
離散均勻分佈 (discrete uniform distribution):每個可能離散數值出現嘅機率都一樣,概率質量函數 係[2] :
f
(
x
)
=
1
n
{\displaystyle f(x)={\frac {1}{n}}}
,當中
n
{\displaystyle n}
係
X
{\displaystyle X}
有幾多個可能數值。
伯努利分佈 (Bernoulli distribution):描述嘅變數
k
{\displaystyle k}
得兩個可能數值,數值係 1 嘅機會率係
p
{\displaystyle p}
,數值係 0 嘅機會率係
q
=
(
1
−
p
)
{\displaystyle q=(1-p)}
,概率質量函數
f
(
k
;
p
)
{\displaystyle f(k;p)}
係[3] :
f
(
k
;
p
)
=
{
p
if
k
=
1
,
q
=
1
−
p
if
k
=
0.
{\displaystyle f(k;p)={\begin{cases}p&{\text{if }}k=1,\\q=1-p&{\text{if }}k=0.\end{cases}}}
廣義伯努利分佈 (generalized Bernoulli distribution / multinoulli distribution):描述嘅變數
k
{\displaystyle k}
有
n
{\displaystyle n}
個離散可能數值,概率質量函數 係[4] :
f
(
i
)
=
{
p
1
if
i
=
1
,
p
2
if
i
=
2
,
p
3
if
i
=
3
,
.
.
.
{\displaystyle f(i)={\begin{cases}p_{1}&{\text{if }}i=1,\\p_{2}&{\text{if }}i=2,\\p_{3}&{\text{if }}i=3,\\...\end{cases}}}
二項分佈 (binomial distribution):描述
n
{\displaystyle n}
次結果二元嘅試驗;想像有個結果係二元-得兩個可能結果(1 同 0)-嘅試驗,例如掟銀仔 ,做
n
{\displaystyle n}
咁多次,每次試驗嘅結果都有
p
{\displaystyle p}
咁多機會率係 1,
q
=
(
1
−
p
)
{\displaystyle q=(1-p)}
咁多機會率係 0,而每次試驗嘅結果都係獨立 嘅(一次試驗嘅結果唔受其他試驗嘅結果影響)。概率質量函數
f
(
k
,
n
,
p
)
{\displaystyle f(k,n,p)}
,即係得出
k
{\displaystyle k}
咁多個 1 嘅機會率係[3] :
f
(
k
,
n
,
p
)
=
Pr
(
k
;
n
,
p
)
=
Pr
(
X
=
k
)
=
(
n
k
)
p
k
(
1
−
p
)
n
−
k
{\displaystyle f(k,n,p)=\Pr(k;n,p)=\Pr(X=k)={\binom {n}{k}}p^{k}(1-p)^{n-k}}
多項分佈 (multinomial distribution):係二項分佈嘅廣義化 ,描述嘅試驗有
k
{\displaystyle k}
個可能結果,做
n
{\displaystyle n}
咁多次(想像掟一粒
k
{\displaystyle k}
面嘅骰仔 掟
n
{\displaystyle n}
咁多次)。概率質量函數 係[5] :
f
(
k
,
n
,
p
)
=
n
!
x
1
!
⋯
x
k
!
p
1
x
1
⋯
p
k
x
k
{\displaystyle f(k,n,p)={\frac {n!}{x_{1}!\cdots x_{k}!}}p_{1}^{x_{1}}\cdots p_{k}^{x_{k}}}
一個二項分佈嘅概率質量函數圖;X 軸係
k
{\displaystyle k}
。
幾何分佈 (geometric distribution):可以指兩個唔同嘅概率分佈,兩者都涉及一個結果二元嘅試驗[6] :
做咗個試驗
k
{\displaystyle k}
次,終於得到 1 次陽性結果,而之前嗰啲試驗結果冚唪唥都係陰性:
Pr
(
X
=
k
)
=
(
1
−
p
)
k
−
1
p
{\displaystyle \Pr(X=k)=(1-p)^{k-1}p}
for
k
=
1
,
2
,
3
,
.
.
.
{\displaystyle {\text{for }}k=1,2,3,...}
k
{\displaystyle k}
代表要做幾多次陰性試驗,先可以得到一次陽性結果:
Pr
(
Y
=
k
)
=
(
1
−
p
)
k
p
{\displaystyle \Pr(Y=k)=(1-p)^{k}p}
for
k
=
0
,
1
,
2
,
3
,
.
.
.
{\displaystyle {\text{for }}k=0,1,2,3,...}
兩個幾何分佈嘅概率質量函數圖;X 軸係
k
{\displaystyle k}
。
撥桑分佈 (Poisson distribution):模擬嘅事件有已知嘅平均發生率,而每件事件嘅發生彼此之間獨立 ,發生嘅次數設做
k
{\displaystyle k}
,概率質量函數 係[7] :
f
(
k
;
λ
)
=
Pr
(
X
=
k
)
=
λ
k
e
−
λ
k
!
{\displaystyle \!f(k;\lambda )=\Pr(X=k)={\frac {\lambda ^{k}e^{-\lambda }}{k!}}}
,當中
λ
{\displaystyle \lambda }
係預期會發生嘅次數(唔一定係整數 )。
撥桑分佈嘅概率質量函數畫做圖嘅樣
p
{\displaystyle p}
係二項分佈當中有嘅一個參數 。
Ash, Robert B. (2008). Basic probability theory (Dover ed.). Mineola, N.Y.: Dover Publications. pp. 66–69.
1941-, Çınlar, E. (Erhan) (2011). Probability and stochastics . New York: Springer. p. 51.
Bertsekas, Dimitri P. (2002). Introduction to Probability . Tsitsiklis, John N., Τσιτσικλής, Γιάννης Ν. Belmont, Mass.: Athena Scientific.
Murphy, K. P. (2012). Machine learning: a probabilistic perspective , p. 35. MIT press.
Ostrovski, Vladimir (May 2017). "Testing equivalence of multinomial distributions". Statistics & Probability Letters . 124: 77–82.
Gallager, R.; van Voorhis, D. (March 1975). "Optimal source codes for geometrically distributed integer alphabets (Corresp.)". IEEE Transactions on Information Theory . 21 (2): 228–230.
Haight, Frank A. (1967), Handbook of the Poisson Distribution , New York, NY, USA: John Wiley & Sons.
Johnson, Norman L.; Kotz, Samuel; Balakrishnan, N. (1994), "14: Lognormal Distributions", Continuous univariate distributions . Vol. 1, Wiley Series in Probability and Mathematical Statistics: Applied Probability and Statistics (2nd ed.), New York: John Wiley & Sons.
Reed, William J.; et al. (2004). "The Double Pareto-Lognormal Distribution – A New Parametric Model for Size Distributions". Communications in Statistics – Theory and Methods . 33 (8): 1733–53.
VAN MONTFORT, M.A.J. (1986). "The Generalized Pareto distribution applied to rainfall depths". Hydrological Sciences Journal . 31 (2): 151–162.
Elfessi, Abdulaziz; Reineke, David M. (2001). "A Bayesian Look at Classical Estimation: The Exponential Distribution". Journal of Statistics Education . 9 (1).
Manikandan, S (1 January 2011). "Frequency distribution". Journal of Pharmacology & Pharmacotherapeutics . 2 (1): 54–55.
Deisenroth,Faisal,Ong, Marc Peter,A Aldo, Cheng Soon (2019). Mathematics for Machine Learning . Cambridge University Press. p. 181.
Ali, Mir M. (1980). "Characterization of the Normal Distribution Among the Continuous Symmetric Spherical Class". Journal of the Royal Statistical Society . Series B (Methodological). 42 (2): 162–164.
Spanos, Aris (1999). Probability Theory and Statistical Inference . New York: Cambridge University Press. pp. 109–130.
MacGillivray, HL (1992). "Shape properties of the g- and h- and Johnson families". Communications in Statistics - Theory and Methods . 21: 1244–1250.
Altman, Douglas G; Bland, J Martin (2005-10-15). "Standard deviations and standard errors". BMJ: British Medical Journal . 331 (7521): 903.
Hazewinkel, Michiel, ed. (2001) [1994], "Joint distribution", Encyclopedia of Mathematics , Springer Science+Business Media B.V. / Kluwer Academic Publishers.
Dinov, Ivo; Christou, Nicolas; Sanchez, Juana (2008). "Central Limit Theorem: New SOCR Applet and Demonstration Activity". Journal of Statistics Education . ASA. 16 (2).