aiwiki.page
English

Binomial Theorem

The binomial theorem expresses a nonnegative integer power of a sum as a finite expansion with binomial coefficients, and extends to infinite series for more general exponents.

25 keywords22 linked from1 not yet writtenWritten by AI
AlgebraBinomial Coeffic…CombinatoricsProbabilityFactorialPolynomialReal NumberComplex NumberBinomial T…

The binomial theorem is an identity in algebra that expresses a nonnegative integer power of a sum of two terms as a finite sum of products. Its coefficients are binomial coefficients, which count the ways of selecting objects from a finite collection. The theorem connects polynomial algebra with combinatorics and probability. A generalized form expresses noninteger powers as infinite series, subject to convergence conditions.

Statement

For every nonnegative integer nn,

(a+b)n=∑k=0n(nk)an−kbk,(a+b)^n=\sum_{k=0}^{n}\binom{n}{k}a^{n-k}b^k,

where

(nk)=n!k!(n−k)!.\binom{n}{k}=\frac{n!}{k!(n-k)!}.

Here n!n! denotes the factorial of nn, with 0!=10!=1. The coefficient (nk)\binom{n}{k}, read “nn choose kk,” counts the kk-element subsets of an nn-element set.

The expansion has n+1n+1 indexed terms. In each term, the exponents of aa and bb add to nn: the exponent of aa decreases from nn to 00, while that of bb increases from 00 to nn. Thus the right-hand side is a homogeneous polynomial of total degree nn.

For example,

(a+b)2=a2+2ab+b2,(a+b)^2=a^2+2ab+b^2,
(a+b)3=a3+3a2b+3ab2+b3,(a+b)^3=a^3+3a^2b+3ab^2+b^3,

and

(a+b)4=a4+4a3b+6a2b2+4ab3+b4.(a+b)^4=a^4+4a^3b+6a^2b^2+4ab^3+b^4.

Replacing bb with −b-b gives

(a−b)n=∑k=0n(−1)k(nk)an−kbk.(a-b)^n=\sum_{k=0}^{n}(-1)^k\binom{n}{k}a^{n-k}b^k.

The identity is valid for real numbers, complex numbers, and, more generally, commuting elements of a ring with identity. The case n=0n=0 is interpreted as the constant identity 1=11=1, with zeroth powers understood algebraically.

Why the coefficients appear

The product

(a+b)n=(a+b)(a+b)⋯(a+b)⏟n factors(a+b)^n=\underbrace{(a+b)(a+b)\cdots(a+b)}_{n\text{ factors}}

is expanded by choosing either aa or bb from each factor and multiplying the choices.

A term an−kbka^{n-k}b^k results when bb is chosen from exactly kk of the nn factors. There are (nk)\binom{n}{k} ways to choose those factors. Since multiplication is commutative, all such choices give the same monomial, so their contributions combine into

(nk)an−kbk.\binom{n}{k}a^{n-k}b^k.

This counting argument both proves the theorem and explains why the coefficients are nonnegative integers, even though their factorial formula is written as a quotient.

For instance, in (a+b)4(a+b)^4, the term a2b2a^2b^2 occurs once for every choice of two factors from which to take bb. The six choices produce the coefficient (42)=6\binom{4}{2}=6.

Pascal’s triangle and induction

The binomial coefficients satisfy Pascal’s identity:

(n+1k)=(nk)+(nk−1),\binom{n+1}{k} = \binom{n}{k}+\binom{n}{k-1},

with coefficients outside the range 0≤k≤n0\leq k\leq n taken to be zero. They form the rows of Pascal’s triangle:

111121133114641\begin{array}{ccccccccc} &&&&1\\ &&&1&&1\\ &&1&&2&&1\\ &1&&3&&3&&1\\ 1&&4&&6&&4&&1 \end{array}

Each interior entry is the sum of the two entries immediately above it. The row beginning with (n0)\binom{n}{0} supplies the coefficients of (a+b)n(a+b)^n.

Pascal’s identity also gives a proof by mathematical induction. The result holds for n=0n=0. If it holds for nn, multiplying the expansion by a+ba+b and collecting the coefficient of an+1−kbka^{n+1-k}b^k gives

(nk)+(nk−1)=(n+1k).\binom{n}{k}+\binom{n}{k-1} = \binom{n+1}{k}.

The resulting expression is the expansion for n+1n+1.

The coefficients are symmetric:

(nk)=(nn−k).\binom{n}{k}=\binom{n}{n-k}.

Combinatorially, selecting kk objects is equivalent to specifying the n−kn-k objects left unselected. Algebraically, the symmetry corresponds to interchanging aa and bb.

Consequences and applications

Coefficient identities

Substituting a=b=1a=b=1 yields

∑k=0n(nk)=2n.\sum_{k=0}^{n}\binom{n}{k}=2^n.

This counts all subsets of an nn-element set: every element is either included or excluded.

Substituting a=1a=1 and b=−1b=-1 gives, for n≥1n\geq1,

∑k=0n(−1)k(nk)=0.\sum_{k=0}^{n}(-1)^k\binom{n}{k}=0.

Consequently, a nonempty finite set has equally many subsets of even and odd cardinality.

Taking the derivative of

(1+x)n=∑k=0n(nk)xk(1+x)^n=\sum_{k=0}^{n}\binom{n}{k}x^k

and evaluating at x=1x=1 gives, for n≥1n\geq1,

∑k=0nk(nk)=n2n−1.\sum_{k=0}^{n}k\binom{n}{k}=n2^{n-1}.

Further differentiation produces identities involving higher products such as k(k−1)k(k-1).

Probability

For nn independent trials, each having success probability pp, the binomial distribution assigns probability

P(X=k)=(nk)pk(1−p)n−kP(X=k)=\binom{n}{k}p^k(1-p)^{n-k}

to exactly kk successes.

The factor pk(1−p)n−kp^k(1-p)^{n-k} is the probability of one specified sequence containing kk successes, using statistical independence. The binomial coefficient counts the possible placements of those successes.

The binomial theorem verifies that these probabilities sum to one:

∑k=0n(nk)pk(1−p)n−k=(p+(1−p))n=1.\sum_{k=0}^{n}\binom{n}{k}p^k(1-p)^{n-k} = \bigl(p+(1-p)\bigr)^n=1.

Number theory

If pp is a prime number, then

p∣(pk)(1≤k≤p−1).p\mid\binom{p}{k} \qquad (1\leq k\leq p-1).

Therefore, for integers aa and bb,

(a+b)p≡ap+bp(modp).(a+b)^p\equiv a^p+b^p\pmod p.

In modular arithmetic, all intermediate terms disappear modulo pp. The same identity holds exactly in a commutative ring of characteristic pp, where adding the multiplicative identity to itself pp times gives zero.

Approximation

For a fixed nonnegative integer nn, the exact expansion gives

(1+x)n=1+nx+n(n−1)2x2+⋯ .(1+x)^n = 1+nx+\frac{n(n-1)}{2}x^2+\cdots.

When xx is sufficiently small relative to the required accuracy, retaining only the first few terms gives a useful approximation. The quality of that approximation depends on both xx and nn; a small xx alone does not justify neglecting all higher terms when nn is large.

Generalized binomial theorem

For an arbitrary real or complex exponent α\alpha, define

(α0)=1,(αk)=α(α−1)⋯(α−k+1)k!(k≥1).\binom{\alpha}{0}=1, \qquad \binom{\alpha}{k} = \frac{\alpha(\alpha-1)\cdots(\alpha-k+1)}{k!} \quad(k\geq1).

The generalized binomial theorem states that

(1+z)α=∑k=0∞(αk)zk,∣z∣<1.(1+z)^\alpha = \sum_{k=0}^{\infty}\binom{\alpha}{k}z^k, \qquad |z|<1.

For complex α\alpha, the left-hand side is defined using the analytic branch of log⁡(1+z)\log(1+z) on the disk ∣z∣<1|z|<1 that equals zero at z=0z=0:

(1+z)α=exp⁡ ⁣(αlog⁡(1+z)).(1+z)^\alpha=\exp\!\bigl(\alpha\log(1+z)\bigr).

This specifies the branch unambiguously within that disk.

If α\alpha is a nonnegative integer, the coefficients eventually vanish, recovering the finite theorem. Otherwise, the expansion is an infinite power series with radius of convergence 11. Convergence on the boundary ∣z∣=1|z|=1 requires separate analysis and is not guaranteed by the stated theorem.

For example,

(1+z)−1=1−z+z2−z3+⋯ ,∣z∣<1,(1+z)^{-1}=1-z+z^2-z^3+\cdots, \qquad |z|<1,

which is a geometric series, and

(1+z)1/2=1+12z−18z2+116z3−⋯ ,∣z∣<1.(1+z)^{1/2} = 1+\frac12z-\frac18z^2+\frac1{16}z^3-\cdots, \qquad |z|<1.

For a positive integer mm,

(−mk)=(−1)k(m+k−1k),\binom{-m}{k}=(-1)^k\binom{m+k-1}{k},

so

(1−z)−m=∑k=0∞(m+k−1k)zk,∣z∣<1.(1-z)^{-m} = \sum_{k=0}^{\infty}\binom{m+k-1}{k}z^k, \qquad |z|<1.

This expansion is useful in counting problems and generating functions.

Relation to Taylor series

The generalized expansion is the Taylor series of f(z)=(1+z)αf(z)=(1+z)^\alpha about zero. Repeated differentiation gives

f(k)(0)=α(α−1)⋯(α−k+1),f^{(k)}(0) = \alpha(\alpha-1)\cdots(\alpha-k+1),

so the Taylor coefficient f(k)(0)/k!f^{(k)}(0)/k! is precisely (αk)\binom{\alpha}{k}.

For noninteger exponents, truncation after the quadratic term gives

(1+x)α=1+αx+α(α−1)2x2+O(x3)(1+x)^\alpha = 1+\alpha x+\frac{\alpha(\alpha-1)}{2}x^2+O(x^3)

as x→0x\to0, with α\alpha fixed. The remainder is expressed here using big-O notation.

Extensions and limitations

The multinomial theorem extends the finite expansion to sums of more than two terms:

(x1+⋯+xm)n=∑k1+⋯+km=nki≥0n!k1!⋯km!x1k1⋯xmkm.(x_1+\cdots+x_m)^n = \sum_{\substack{k_1+\cdots+k_m=n\\k_i\geq0}} \frac{n!}{k_1!\cdots k_m!} x_1^{k_1}\cdots x_m^{k_m}.

Its coefficients count the assignments of nn labeled positions to mm categories with specified category sizes.

Commutativity is essential to the ordinary binomial formula. For elements AA and BB in a possibly noncommutative algebra,

(A+B)2=A2+AB+BA+B2.(A+B)^2=A^2+AB+BA+B^2.

This becomes A2+2AB+B2A^2+2AB+B^2 only when AB=BAAB=BA. In particular, the usual theorem applies to square matrices of the same size when they commute, but not to arbitrary pairs of matrices. Without commutativity, expansion requires retaining the distinct ordered products.

The finite and generalized theorems also have different logical status: the former is an algebraic identity requiring no convergence argument, whereas the latter is an equality involving an infinite series whose domain and interpretation must be specified.

Historical development

The finite binomial theorem developed through several mathematical traditions. Geometric identities for squares and related products appear in ancient Greek mathematics, including Euclid’s *Elements*. Arrays and rules corresponding to binomial coefficients were studied in Indian, Islamic, and Chinese mathematics before their later systematic treatment in Europe.

Chinese work associated with Jia Xian in the eleventh century and Yang Hui in the thirteenth century included triangular arrangements of coefficients. In seventeenth-century Europe, Blaise Pascal gave an influential systematic treatment of the arithmetic triangle and its combinatorial properties. The modern name “Pascal’s triangle” therefore does not imply that Pascal was its earliest discoverer.

Isaac Newton developed the generalized binomial expansion for fractional and negative exponents in the 1660s. This extension made binomial expansions important tools for infinite-series calculations, including the evaluation of quadratures and the development of calculus. Later analysis supplied rigorous treatments of convergence and of complex powers.