Clever Geek Handbook
📜 ⬆️ ⬇️

Multinomial distribution

The multinomial (polynomial) distribution in probability theory is a generalization of the binomial distribution for the case of n> 1 independent trials of a random experiment with k> 2 possible outcomes.

Definition

Let beXone,...,Xn {\ displaystyle X_ {1}, \ ldots, X_ {n}}   - independent identically distributed random variables , such that their distribution is given by a probability function :

P(Xi=j)=pj,j=one,...,k{\ displaystyle \ mathbb {P} (X_ {i} = j) = p_ {j}, \; j = 1, \ ldots, k}   .

Intuitive event{Xi=j} {\ displaystyle \ {X_ {i} = j \}}   means test numberi {\ displaystyle i}   led to the outcomej {\ displaystyle j}   . Let the random variableYj {\ displaystyle Y_ {j}}   equal to the number of trials that led to the outcomej {\ displaystyle j}   :

Yj=Σi=onenone{Xi=j},j=one,...,k{\ displaystyle Y_ {j} = \ sum _ {i = 1} ^ {n} \ mathbf {1} _ {\ {X_ {i} = j \}}, \; j = 1, \ ldots, k}   .

Then the distribution of the vectorY=(Yone,...,Yk)⊤ {\ displaystyle \ mathbf {Y} = (Y_ {1}, \ ldots, Y_ {k}) ^ {\ top}}   has a probability function

pY(y)={(nyone...yk)poneyone...pkyk,Σj=onekyj=n0,Σj=onekyj≠n,y=(yone,...,yk)⊤∈Nonek{\ displaystyle p _ {\ mathbf {Y}} (\ mathbf {y}) = \ left \ {{\ begin {matrix} {n \ choose {y_ {1} \ ldots y_ {k}}} p_ {1} ^ {y_ {1}} \ ldots p_ {k} ^ {y_ {k}}, & \ sum \ limits _ {j = 1} ^ {k} y_ {j} = n \\ 0, & \ sum \ limits _ {j = 1} ^ {k} y_ {j} \ not = n \ end {matrix}} \ right., \ quad \ mathbf {y} = (y_ {1}, \ ldots, y_ {k} ) ^ {\ top} \ in \ mathbb {N} _ {1} ^ {k}}   ,

Where

(nyone...yk)≡n!yone!...yk!{\ displaystyle {n \ choose {y_ {1} \ ldots y_ {k}}} \ equiv {\ frac {n!} {y_ {1}! \ ldots y_ {k}!}}}   - multinomial coefficient .

Medium vector and covariance matrix

Mathematical expectation of a random variableYj {\ displaystyle Y_ {j}}   has the form:E[Yj]=npj {\ displaystyle \ mathbb {E} [Y_ {j}] = np_ {j}}   . Diagonal elements of the covariance matrixΣ=(σij) {\ displaystyle \ Sigma = (\ sigma _ {ij})}   are variances of binomial random variables, and therefore

σjj=D[Yj]=npj(one-pj),j=one,...,k{\ displaystyle \ sigma _ {jj} = \ mathrm {D} [Y_ {j}] = np_ {j} (1-p_ {j}), \; j = 1, \ ldots, k}   .

For the remaining items we have

σij=cov(Yi,Yj)=-npipj,i≠j{\ displaystyle \ sigma _ {ij} = \ mathrm {cov} (Y_ {i}, Y_ {j}) = - np_ {i} p_ {j}, \; i \ not = j}   .

The rank of the covariance matrix of the multinomial distribution isk-one {\ displaystyle k-1}   .

Source - https://ru.wikipedia.org/w/index.php?title=Multinomial_distribution&oldid=96171670


More articles:

  • Curved Sedge
  • NBA 1999 Draft
  • Antonovich, Tatyana Mikhailovna
  • Leshchina, Sergey Ivanovich
  • Björling, Jussi
  • Bosconero
  • Novaleza
  • Bojnice
  • San Carlo Canavese
  • Torre Pellice

All articles

Clever Geek | 2019