{
  "id": "binomial-theorem",
  "title": "Binomial Theorem",
  "lang": "en",
  "source_lang": "en",
  "summary": "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.",
  "category": "mathematics",
  "subcategory": "algebra",
  "type": "concept",
  "aliases": [
    "Binomial expansion"
  ],
  "generated_at": "2026-11-04",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/binomial-theorem/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/binomial-theorem/",
      "json": "https://aiwiki.page/zh/binomial-theorem.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "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."
    },
    {
      "heading": "Statement",
      "text": "For every nonnegative integer \\(n\\),\n\n\\[(a+b)^n=\\sum_{k=0}^{n}\\binom{n}{k}a^{n-k}b^k,\\]\n\nwhere\n\n\\[\\binom{n}{k}=\\frac{n!}{k!(n-k)!}.\\]\n\nHere \\(n!\\) denotes the factorial of \\(n\\), with \\(0!=1\\). The coefficient \\(\\binom{n}{k}\\), read “\\(n\\) choose \\(k\\),” counts the \\(k\\)-element subsets of an \\(n\\)-element set.\n\nThe expansion has \\(n+1\\) indexed terms. In each term, the exponents of \\(a\\) and \\(b\\) add to \\(n\\): the exponent of \\(a\\) decreases from \\(n\\) to \\(0\\), while that of \\(b\\) increases from \\(0\\) to \\(n\\). Thus the right-hand side is a homogeneous polynomial of total degree \\(n\\).\n\nFor example,\n\n\\[(a+b)^2=a^2+2ab+b^2,\\]\n\n\\[(a+b)^3=a^3+3a^2b+3ab^2+b^3,\\]\n\nand\n\n\\[(a+b)^4=a^4+4a^3b+6a^2b^2+4ab^3+b^4.\\]\n\nReplacing \\(b\\) with \\(-b\\) gives\n\n\\[(a-b)^n=\\sum_{k=0}^{n}(-1)^k\\binom{n}{k}a^{n-k}b^k.\\]\n\nThe identity is valid for real numbers, complex numbers, and, more generally, commuting elements of a ring with identity. The case \\(n=0\\) is interpreted as the constant identity \\(1=1\\), with zeroth powers understood algebraically."
    },
    {
      "heading": "Why the coefficients appear",
      "text": "The product\n\n\\[(a+b)^n=\\underbrace{(a+b)(a+b)\\cdots(a+b)}_{n\\text{ factors}}\\]\n\nis expanded by choosing either \\(a\\) or \\(b\\) from each factor and multiplying the choices.\n\nA term \\(a^{n-k}b^k\\) results when \\(b\\) is chosen from exactly \\(k\\) of the \\(n\\) factors. There are \\(\\binom{n}{k}\\) ways to choose those factors. Since multiplication is commutative, all such choices give the same monomial, so their contributions combine into\n\n\\[\\binom{n}{k}a^{n-k}b^k.\\]\n\nThis counting argument both proves the theorem and explains why the coefficients are nonnegative integers, even though their factorial formula is written as a quotient.\n\nFor instance, in \\((a+b)^4\\), the term \\(a^2b^2\\) occurs once for every choice of two factors from which to take \\(b\\). The six choices produce the coefficient \\(\\binom{4}{2}=6\\)."
    },
    {
      "heading": "Pascal’s triangle and induction",
      "text": "The binomial coefficients satisfy Pascal’s identity:\n\n\\[\\binom{n+1}{k} = \\binom{n}{k}+\\binom{n}{k-1},\\]\n\nwith coefficients outside the range \\(0\\leq k\\leq n\\) taken to be zero. They form the rows of Pascal’s triangle:\n\n\\[\\begin{array}{ccccccccc} &&&&1\\\\ &&&1&&1\\\\ &&1&&2&&1\\\\ &1&&3&&3&&1\\\\ 1&&4&&6&&4&&1 \\end{array}\\]\n\nEach interior entry is the sum of the two entries immediately above it. The row beginning with \\(\\binom{n}{0}\\) supplies the coefficients of \\((a+b)^n\\).\n\nPascal’s identity also gives a proof by mathematical induction. The result holds for \\(n=0\\). If it holds for \\(n\\), multiplying the expansion by \\(a+b\\) and collecting the coefficient of \\(a^{n+1-k}b^k\\) gives\n\n\\[\\binom{n}{k}+\\binom{n}{k-1} = \\binom{n+1}{k}.\\]\n\nThe resulting expression is the expansion for \\(n+1\\).\n\nThe coefficients are symmetric:\n\n\\[\\binom{n}{k}=\\binom{n}{n-k}.\\]\n\nCombinatorially, selecting \\(k\\) objects is equivalent to specifying the \\(n-k\\) objects left unselected. Algebraically, the symmetry corresponds to interchanging \\(a\\) and \\(b\\)."
    },
    {
      "heading": "Consequences and applications",
      "text": "Coefficient identities\n\nSubstituting \\(a=b=1\\) yields\n\n\\[\\sum_{k=0}^{n}\\binom{n}{k}=2^n.\\]\n\nThis counts all subsets of an \\(n\\)-element set: every element is either included or excluded.\n\nSubstituting \\(a=1\\) and \\(b=-1\\) gives, for \\(n\\geq1\\),\n\n\\[\\sum_{k=0}^{n}(-1)^k\\binom{n}{k}=0.\\]\n\nConsequently, a nonempty finite set has equally many subsets of even and odd cardinality.\n\nTaking the derivative of\n\n\\[(1+x)^n=\\sum_{k=0}^{n}\\binom{n}{k}x^k\\]\n\nand evaluating at \\(x=1\\) gives, for \\(n\\geq1\\),\n\n\\[\\sum_{k=0}^{n}k\\binom{n}{k}=n2^{n-1}.\\]\n\nFurther differentiation produces identities involving higher products such as \\(k(k-1)\\).\n\nProbability\n\nFor \\(n\\) independent trials, each having success probability \\(p\\), the binomial distribution assigns probability\n\n\\[P(X=k)=\\binom{n}{k}p^k(1-p)^{n-k}\\]\n\nto exactly \\(k\\) successes.\n\nThe factor \\(p^k(1-p)^{n-k}\\) is the probability of one specified sequence containing \\(k\\) successes, using statistical independence. The binomial coefficient counts the possible placements of those successes.\n\nThe binomial theorem verifies that these probabilities sum to one:\n\n\\[\\sum_{k=0}^{n}\\binom{n}{k}p^k(1-p)^{n-k} = \\bigl(p+(1-p)\\bigr)^n=1.\\]\n\nNumber theory\n\nIf \\(p\\) is a prime number, then\n\n\\[p\\mid\\binom{p}{k} \\qquad (1\\leq k\\leq p-1).\\]\n\nTherefore, for integers \\(a\\) and \\(b\\),\n\n\\[(a+b)^p\\equiv a^p+b^p\\pmod p.\\]\n\nIn modular arithmetic, all intermediate terms disappear modulo \\(p\\). The same identity holds exactly in a commutative ring of characteristic \\(p\\), where adding the multiplicative identity to itself \\(p\\) times gives zero.\n\nApproximation\n\nFor a fixed nonnegative integer \\(n\\), the exact expansion gives\n\n\\[(1+x)^n = 1+nx+\\frac{n(n-1)}{2}x^2+\\cdots.\\]\n\nWhen \\(x\\) 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 \\(x\\) and \\(n\\); a small \\(x\\) alone does not justify neglecting all higher terms when \\(n\\) is large."
    },
    {
      "heading": "Generalized binomial theorem",
      "text": "For an arbitrary real or complex exponent \\(\\alpha\\), define\n\n\\[\\binom{\\alpha}{0}=1, \\qquad \\binom{\\alpha}{k} = \\frac{\\alpha(\\alpha-1)\\cdots(\\alpha-k+1)}{k!} \\quad(k\\geq1).\\]\n\nThe generalized binomial theorem states that\n\n\\[(1+z)^\\alpha = \\sum_{k=0}^{\\infty}\\binom{\\alpha}{k}z^k, \\qquad |z|<1.\\]\n\nFor complex \\(\\alpha\\), the left-hand side is defined using the analytic branch of \\(\\log(1+z)\\) on the disk \\(|z|<1\\) that equals zero at \\(z=0\\):\n\n\\[(1+z)^\\alpha=\\exp\\!\\bigl(\\alpha\\log(1+z)\\bigr).\\]\n\nThis specifies the branch unambiguously within that disk.\n\nIf \\(\\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 \\(1\\). Convergence on the boundary \\(|z|=1\\) requires separate analysis and is not guaranteed by the stated theorem.\n\nFor example,\n\n\\[(1+z)^{-1}=1-z+z^2-z^3+\\cdots, \\qquad |z|<1,\\]\n\nwhich is a geometric series, and\n\n\\[(1+z)^{1/2} = 1+\\frac12z-\\frac18z^2+\\frac1{16}z^3-\\cdots, \\qquad |z|<1.\\]\n\nFor a positive integer \\(m\\),\n\n\\[\\binom{-m}{k}=(-1)^k\\binom{m+k-1}{k},\\]\n\nso\n\n\\[(1-z)^{-m} = \\sum_{k=0}^{\\infty}\\binom{m+k-1}{k}z^k, \\qquad |z|<1.\\]\n\nThis expansion is useful in counting problems and generating functions.\n\nRelation to Taylor series\n\nThe generalized expansion is the Taylor series of \\(f(z)=(1+z)^\\alpha\\) about zero. Repeated differentiation gives\n\n\\[f^{(k)}(0) = \\alpha(\\alpha-1)\\cdots(\\alpha-k+1),\\]\n\nso the Taylor coefficient \\(f^{(k)}(0)/k!\\) is precisely \\(\\binom{\\alpha}{k}\\).\n\nFor noninteger exponents, truncation after the quadratic term gives\n\n\\[(1+x)^\\alpha = 1+\\alpha x+\\frac{\\alpha(\\alpha-1)}{2}x^2+O(x^3)\\]\n\nas \\(x\\to0\\), with \\(\\alpha\\) fixed. The remainder is expressed here using big-O notation."
    },
    {
      "heading": "Extensions and limitations",
      "text": "The multinomial theorem extends the finite expansion to sums of more than two terms:\n\n\\[(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}.\\]\n\nIts coefficients count the assignments of \\(n\\) labeled positions to \\(m\\) categories with specified category sizes.\n\nCommutativity is essential to the ordinary binomial formula. For elements \\(A\\) and \\(B\\) in a possibly noncommutative algebra,\n\n\\[(A+B)^2=A^2+AB+BA+B^2.\\]\n\nThis becomes \\(A^2+2AB+B^2\\) only when \\(AB=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.\n\nThe 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."
    },
    {
      "heading": "Historical development",
      "text": "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.\n\nChinese 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.\n\nIsaac 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."
    }
  ],
  "links": [
    "algebra",
    "binomial-coefficient",
    "combinatorics",
    "probability",
    "factorial",
    "polynomial",
    "real-number",
    "complex-number",
    "ring-mathematics",
    "pascals-triangle",
    "mathematical-induction",
    "derivative",
    "binomial-distribution",
    "statistical-independence",
    "prime-number",
    "modular-arithmetic",
    "power-series",
    "geometric-series",
    "generating-function",
    "taylor-series",
    "big-o-notation",
    "matrix",
    "euclids-elements",
    "blaise-pascal"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}