Probability Generating Functions in IB Mathematics | IB 数学：概率生成函数

📚 Probability Generating Functions in IB Mathematics | IB 数学：概率生成函数

Probability Generating Functions (PGFs) are a powerful tool in probability theory, particularly useful in the IB Mathematics Analysis and Approaches Higher Level curriculum. They provide an elegant algebraic method for handling discrete random variables and their distributions.

概率生成函数（PGF）是概率论中一个强大的工具，尤其适用于IB数学分析与方法高级水平课程。它为处理离散随机变量及其分布提供了一种优雅的代数方法。

1. Definition | 定义

Let X be a discrete random variable taking non-negative integer values 0, 1, 2, … The probability generating function of X is defined as the expected value of t^X, written as G_X(t) = E(t^X).

设X是一个取非负整数值（0, 1, 2, …）的离散随机变量。X的概率生成函数定义为t^X的期望值，记为G_X(t) = E(t^X)。

Expanding this definition explicitly, if P(X = x) = p_x, then the PGF can be written as:

如果P(X = x) = p_x，则概率生成函数可以显式写为：

G_X(t) = Σ pₓ tˣ = p₀ + p₁t + p₂t² + p₃t³ + …

where the sum is taken over all possible values x = 0, 1, 2, … and t is a dummy variable (usually |t| ≤ 1 for convergence).

其中求和对x的所有可能取值（x = 0, 1, 2, …）进行，t是一个哑变量（通常|t| ≤ 1以保证收敛性）。

The PGF is a power series representation of the probability mass function. Since the probabilities pₓ sum to 1, we note that G_X(1) = Σ pₓ = 1.

概率生成函数是概率质量函数的幂级数表示。由于概率pₓ之和为1，我们注意到G_X(1) = Σ pₓ = 1。

2. Key Properties | 关键性质

The PGF satisfies several fundamental properties that make it useful for calculations in IB Mathematics.

概率生成函数具有几个基本性质，使其在IB数学计算中非常有用。

Property 1: G_X(1) = 1

性质1：G_X(1) = 1

Substituting t = 1 into G_X(t) = Σ pₓ tˣ gives G_X(1) = Σ pₓ = 1, since the total probability over all outcomes must equal 1.

将t = 1代入G_X(t) = Σ pₓ tˣ得G_X(1) = Σ pₓ = 1，因为所有结果的概率总和必须等于1。

Property 2: Derivatives at t = 1

性质2：在t = 1处的导数

The first derivative evaluated at t = 1 gives the mean: G’_X(1) = E(X). This follows from differentiating the power series term by term: G’_X(t) = Σ x·pₓ·tˣ⁻¹, and setting t = 1 yields Σ x·pₓ = E(X).

在t = 1处的一阶导数给出均值：G’_X(1) = E(X)。这通过对幂级数逐项求导得出：G’_X(t) = Σ x·pₓ·tˣ⁻¹，令t = 1即得Σ x·pₓ = E(X)。

Property 3: Variance from second derivative

性质3：由二阶导数求方差

The second derivative at t = 1 gives G”_X(1) = E[X(X-1)]. Therefore, Var(X) = G”_X(1) + G’_X(1) – [G’_X(1)]².

在t = 1处的二阶导数给出G”_X(1) = E[X(X-1)]。因此，Var(X) = G”_X(1) + G’_X(1) – [G’_X(1)]²。

Property 4: Probability extraction

性质4：提取概率

The probability P(X = k) is the coefficient of tᵏ in the power series expansion of G_X(t), i.e., P(X = k) = G_X⁽ᵏ⁾(0) / k!.

概率P(X = k)是G_X(t)幂级数展开中tᵏ的系数，即P(X = k) = G_X⁽ᵏ⁾(0) / k!。

3. PGFs of Standard Distributions | 常见分布的概率生成函数

In the IB syllabus, students are expected to know the PGFs of several standard discrete distributions.

在IB教学大纲中，学生需要掌握几种标准离散分布的概率生成函数。

Bernoulli(p): X ∈ {0, 1} with P(X = 1) = p, P(X = 0) = 1 – p. Then G_X(t) = (1 – p) + pt = 1 – p + pt.
伯努利分布(p)：X ∈ {0, 1}，其中P(X = 1) = p，P(X = 0) = 1 – p。则G_X(t) = (1 – p) + pt = 1 – p + pt。

Binomial(n, p): X ~ B(n, p), the number of successes in n independent Bernoulli trials. Then G_X(t) = (1 – p + pt)ⁿ.
二项分布(n, p)：X ~ B(n, p)，即n次独立伯努利试验中的成功次数。则G_X(t) = (1 – p + pt)ⁿ。

Geometric(p): X denotes the number of trials until the first success. Then G_X(t) = pt / (1 – (1-p)t).
几何分布(p)：X表示直到首次成功所需的试验次数。则G_X(t) = pt / (1 – (1-p)t)。

Poisson(λ): X ~ Po(λ). Then G_X(t) = e^(λ(t-1)) = e^(λt – λ).
泊松分布(λ)：X ~ Po(λ)。则G_X(t) = e^(λ(t-1)) = e^(λt – λ)。

These results can be derived directly from the definitions and are worth memorising for the examination.

这些结果可以从定义直接推导得出，建议在考试前熟记。

4. Sums of Independent Random Variables | 独立随机变量之和

One of the most important applications of PGFs in IB Mathematics is dealing with sums of independent random variables. Let X and Y be independent discrete random variables with PGFs G_X(t) and G_Y(t) respectively. Define Z = X + Y.

在IB数学中，概率生成函数最重要的应用之一是处理独立随机变量的和。设X和Y是相互独立的离散随机变量，其概率生成函数分别为G_X(t)和G_Y(t)。定义Z = X + Y。

Then the PGF of Z is simply the product of the individual PGFs:

则Z的概率生成函数简单地等于各个概率生成函数的乘积：

G_Z(t) = G_X(t) · G_Y(t)

This property extends to more than two variables: the PGF of the sum of n independent random variables is the product of their individual PGFs.

该性质可以推广到两个以上的变量：n个独立随机变量之和的概率生成函数等于它们各自概率生成函数的乘积。

For example, if X₁, X₂, …, Xₙ are independent Bernoulli(p) random variables, then their sum Sₙ = X₁ + X₂ + … + Xₙ has PGF G_S(t) = (1 – p + pt)ⁿ, which confirms that Sₙ ~ B(n, p).

例如，如果X₁, X₂, …, Xₙ是独立的伯努利(p)随机变量，则它们的和Sₙ = X₁ + X₂ + … + Xₙ的概率生成函数为G_S(t) = (1 – p + pt)ⁿ，这证实了Sₙ ~ B(n, p)。

Similarly, if X ~ Po(λ) and Y ~ Po(μ) are independent, then X + Y ~ Po(λ+μ). The PGF method provides an elegant proof: G_(X+Y)(t) = e^(λ(t-1)) · e^(μ(t-1)) = e^((λ+μ)(t-1)), which is the PGF of a Poisson distribution with parameter λ+μ.

类似地，如果X ~ Po(λ)和Y ~ Po(μ)相互独立，则X + Y ~ Po(λ+μ)。概率生成函数方法提供了一个优雅的证明：G_(X+Y)(t) = e^(λ(t-1)) · e^(μ(t-1)) = e^((λ+μ)(t-1))，这正是参数为λ+μ的泊松分布的概率生成函数。

5. Using PGFs for Expectation and Variance | 利用概率生成函数求期望与方差

The PGF provides a systematic way to compute the mean and variance without explicitly summing infinite series.

概率生成函数提供了一种系统化的方法来计算均值和方差，无需显式地求和无穷级数。

Mean: E(X) = G’_X(1)

均值：E(X) = G’_X(1)

E[X(X-1)]: E[X(X-1)] = G”_X(1)

E[X(X-1)]：E[X(X-1)] = G”_X(1)

Variance: Var(X) = E(X²) – [E(X)]² = G”_X(1) + G’_X(1) – [G’_X(1)]²

方差：Var(X) = E(X²) – [E(X)]² = G”_X(1) + G’_X(1) – [G’_X(1)]²

Let us verify this for the Poisson distribution. Given G_X(t) = e^(λ(t-1)), we compute:

让我们以泊松分布为例来验证这一方法。已知G_X(t) = e^(λ(t-1))，我们计算：

G’_X(t) = λe^(λ(t-1)), G’_X(1) = λ

G”_X(t) = λ²e^(λ(t-1)), G”_X(1) = λ²

Therefore E(X) = λ and Var(X) = G”_X(1) + G’_X(1) – [G’_X(1)]² = λ² + λ – λ² = λ. Both the mean and variance of a Poisson distribution equal λ, as expected.

因此E(X) = λ，Var(X) = G”_X(1) + G’_X(1) – [G’_X(1)]² = λ² + λ – λ² = λ。泊松分布的均值和方差都等于λ，与预期一致。

6. Worked Example | 例题分析

Let us work through a typical IB examination question step by step.

让我们逐步解答一道典型的IB考试题目。

Problem: A random variable X has probability generating function G_X(t) = (0.4 + 0.6t)⁵.

题目：随机变量X的概率生成函数为G_X(t) = (0.4 + 0.6t)⁵。

(a) State the distribution of X.

（a）指出X的分布。

(b) Find P(X = 3).

（b）求P(X = 3)。

（c）求E(X)和Var(X)。

Solution (a): Comparing G_X(t) = (1 – p + pt)ⁿ with G_X(t) = (0.4 + 0.6t)⁵, we identify 1 – p = 0.4 and p = 0.6 with n = 5. Hence X ~ B(5, 0.6).

解答（a）：将G_X(t) = (1 – p + pt)ⁿ与G_X(t) = (0.4 + 0.6t)⁵比较，我们识别出1 – p = 0.4，p = 0.6，且n = 5。因此X ~ B(5, 0.6)。

Solution (b): Since X ~ B(5, 0.6), we use the binomial probability formula:

解答（b）：由于X ~ B(5, 0.6)，我们使用二项分布概率公式：

P(X = 3) = C(5,3) × (0.6)³ × (0.4)² = 10 × 0.216 × 0.16 = 0.3456

Solution (c): Using the known formulas for the binomial distribution, or differentiating the PGF:

解答（c）：使用二项分布已知公式，或对概率生成函数求导：

E(X) = np = 5 × 0.6 = 3, Var(X) = np(1-p) = 5 × 0.6 × 0.4 = 1.2

Alternatively, differentiate G_X(t): G’_X(t) = 5(0.4 + 0.6t)⁴ × 0.6, so G’_X(1) = 5 × 1⁴ × 0.6 = 3.

也可以对G_X(t)求导：G’_X(t) = 5(0.4 + 0.6t)⁴ × 0.6，所以G’_X(1) = 5 × 1⁴ × 0.6 = 3。

7. Applications and Exam Tips | 应用与考试技巧

The PGF is a versatile tool with several important applications beyond simple distribution identification.

概率生成函数是一个多用途工具，除了识别分布之外，还有几个重要的应用。

Application 1: Proving distribution properties

应用1：证明分布性质

When two independent Poisson variables are added, the PGF product proves the resulting distribution is Poisson with summed parameters. This is a common examination question asking students to ‘show that’.

当两个独立的泊松变量相加时，概率生成函数的乘积证明了结果分布仍为泊松分布，参数为两者之和。这是常见的考题类型，要求学生”证明”结论。

Application 2: Mixed distributions

应用2：混合分布

Sometimes a variable Z is defined as a compound or mixed random variable. The PGF provides a clear framework for finding the distribution or computing moments.

有时变量Z被定义为复合或混合随机变量。概率生成函数为找到其分布或计算矩提供了清晰的框架。

Application 3: Sums of random number of variables

应用3：随机个随机变量之和

If N is a random variable and X₁, X₂, … are i.i.d. random variables independent of N, then the PGF of S = X₁ + X₂ + … + X_N has a particularly elegant form:

如果N是一个随机变量，X₁, X₂, …是与N独立且同分布的随机变量，则S = X₁ + X₂ + … + X_N的概率生成函数有一个特别优雅的形式：

G_S(t) = G_N(G_X(t))

where G_N is the PGF of N and G_X is the common PGF of each Xᵢ.

其中G_N是N的概率生成函数，G_X是每个Xᵢ共同的概率生成函数。

Exam Tips:

考试技巧：

Always start by checking G_X(1) = 1 as a sanity check.
始终先检查G_X(1) = 1作为合理性检验。

Memorise the PGFs of the four standard distributions covered in the syllabus: Bernoulli, Binomial, Geometric, and Poisson.
熟记教学大纲中四种标准分布的PGF：伯努利、二项、几何和泊松分布。

When computing variance, be careful to use the formula Var(X) = G”_X(1) + G’_X(1) – [G’_X(1)]², not G”_X(1) alone.
计算方差时，注意使用公式Var(X) = G”_X(1) + G’_X(1) – [G’_X(1)]²，而不是仅仅使用G”_X(1)。

For sums of independent variables, write down the product rule clearly and simplify step by step.
对于独立变量之和，清晰地写出乘积规则并逐步化简。

8. Common Pitfalls | 常见错误

Students often make the following mistakes when working with PGFs in IB examinations.

学生在IB考试中使用概率生成函数时常犯以下错误。

Mistake \| 错误	Correction \| 纠正
Writing G_X(t) = E(e^(tX)) instead of E(t^X)	The PGF uses t^X, not e^(tX). The moment generating function uses e^(tX).
把G_X(t)写成E(e^(tX))而非E(t^X)	PGF使用的是t^X，而非e^(tX)。矩生成函数才使用e^(tX)。
Forgetting that G”_X(1) = E[X(X-1)], not E(X²)	E(X²) = G”_X(1) + G’_X(1)
忘记G”_X(1) = E[X(X-1)]，而非E(X²)	E(X²) = G”_X(1) + G’_X(1)
Confusing the PGF of Bernoulli and Binomial	Bernoulli: 1 – p + pt; Binomial: (1 – p + pt)ⁿ
混淆伯努利分布和二项分布的PGF	伯努利：1 – p + pt；二项：(1 – p + pt)ⁿ
Using the PGF for continuous variables without checking the domain	PGFs are only defined for discrete random variables taking non-negative integer values.
未检查定义域就对连续变量使用PGF	PGF仅定义在取非负整数值的离散随机变量上。

9. Connection with Moment Generating Functions | 与矩生成函数的联系

While the PGF uses t^X, the moment generating function (MGF) uses e^(tX). The two are related by the substitution t = e^s in the PGF, giving G_X(e^s) = E((e^s)^X) = E(e^(sX)) = M_X(s).

概率生成函数使用t^X，而矩生成函数（MGF）使用e^(tX)。两者通过PGF中的代换t = e^s联系起来，即G_X(e^s) = E((e^s)^X) = E(e^(sX)) = M_X(s)。

This connection means that techniques developed for PGFs can often be translated to MGFs and vice versa. In IB Mathematics, students are expected to be comfortable with both, recognising when each is more convenient.

这一联系意味着为PGF开发的技术通常可以转化为MGF，反之亦然。在IB数学中，学生需要熟练掌握两者，并认识到何时使用哪一个更方便。

The PGF tends to be simpler for discrete distributions on non-negative integers, while the MGF is more general and applies to both discrete and continuous distributions.

PGF对非负整数上的离散分布更简洁，而MGF更具一般性，适用于离散和连续分布。

10. Practice Questions | 练习题目

Test your understanding with these practice questions.

通过以下练习题目检验你的理解。

Question 1: If X has PGF G_X(t) = (0.3 + 0.7t)⁸, find P(X = 5), E(X), and Var(X).

练习1：若X的PGF为G_X(t) = (0.3 + 0.7t)⁸，求P(X = 5)、E(X)和Var(X)。

Question 2: Let X ~ Po(3) and Y ~ Po(4) be independent. Use PGFs to find the distribution of X + Y, then compute P(X + Y = 5).

练习2：设X ~ Po(3)和Y ~ Po(4)独立。使用PGF求X + Y的分布，然后计算P(X + Y = 5)。

Question 3: A random variable X has PGF G_X(t) = e^(2(t-1)). Identify the distribution parameter and find Var(X).

练习3：随机变量X的PGF为G_X(t) = e^(2(t-1))。识别分布参数并求Var(X)。

Question 4: If X₁, X₂, …, X₁₀ are independent Bernoulli(0.4) variables and S = X₁ + X₂ + … + X₁₀, state the PGF of S and hence find P(S = 6).

练习4：若X₁, X₂, …, X₁₀是独立的伯努利(0.4)变量，且S = X₁ + X₂ + … + X₁₀，写出S的PGF并由此求P(S = 6)。

Solutions:

解答：

Answer 1: X ~ B(8, 0.7). P(X = 5) = C(8,5)(0.7)⁵(0.3)³ ≈ 0.2541. E(X) = 8 × 0.7 = 5.6. Var(X) = 8 × 0.7 × 0.3 = 1.68.

答案1：X ~ B(8, 0.7)。P(X = 5) = C(8,5)(0.7)⁵(0.3)³ ≈ 0.2541。E(X) = 8 × 0.7 = 5.6。Var(X) = 8 × 0.7 × 0.3 = 1.68。

Answer 2: G_(X+Y)(t) = e^(3(t-1)) × e^(4(t-1)) = e^(7(t-1)), so X + Y ~ Po(7). P(X + Y = 5) = (e⁻⁷ × 7⁵) / 5! ≈ 0.1277.

答案2：G_(X+Y)(t) = e^(3(t-1)) × e^(4(t-1)) = e^(7(t-1))，所以X + Y ~ Po(7)。P(X + Y = 5) = (e⁻⁷ × 7⁵) / 5! ≈ 0.1277。

Answer 3: Comparing with e^(λ(t-1)), we have λ = 2. Therefore X ~ Po(2) and Var(X) = 2.

答案3：与e^(λ(t-1))比较，得λ = 2。因此X ~ Po(2)，Var(X) = 2。

Answer 4: G_S(t) = (0.6 + 0.4t)¹⁰, so S ~ B(10, 0.4). P(S = 6) = C(10,6)(0.4)⁶(0.6)⁴ ≈ 0.1115.

答案4：G_S(t) = (0.6 + 0.4t)¹⁰，所以S ~ B(10, 0.4)。P(S = 6) = C(10,6)(0.4)⁶(0.6)⁴ ≈ 0.1115。

11. Summary | 总结

The probability generating function is an indispensable tool in IB Mathematics HL. It provides a unified framework for analysing discrete probability distributions, simplifying complex problems involving sums of independent variables, and offering efficient routes to computing means and variances.

概率生成函数是IB数学高级水平中不可或缺的工具。它为分析离散概率分布提供了统一的框架，简化了涉及独立变量之和的复杂问题，并提供了计算均值和方差的高效途径。

The key ideas to remember are the definition G_X(t) = E(t^X), the moment formulas E(X) = G’_X(1) and Var(X) = G”_X(1) + G’_X(1) – [G’_X(1)]², and the product rule for independent sums G_(X+Y)(t) = G_X(t) · G_Y(t).

需要记住的关键概念包括：定义G_X(t) = E(t^X)，矩公式E(X) = G’_X(1)和Var(X) = G”_X(1) + G’_X(1) – [G’_X(1)]²，以及独立变量之和的乘积规则G_(X+Y)(t) = G_X(t) · G_Y(t)。

Mastery of PGFs will serve you well not only in the IB examination but also in university-level probability and statistics courses. Practice regularly with past paper questions to build fluency and confidence.

熟练掌握概率生成函数不仅对IB考试大有裨益，在大学的概率论与统计课程中也将使你受益良多。建议定期练习历年真题，以提升熟练度和信心。

Published by TutorHao | Mathematics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Probability Generating Functions in IB Mathematics | IB 数学：概率生成函数