📚 Adding Independent Random Variables | 独立随机变量的加法
In probability and statistics, one of the most powerful modelling tools is the ability to combine random variables. When we study independent random variables, we can easily compute the expectation and variance of their sum and identify the distribution of the resulting variable. This topic appears throughout the IB Diploma Mathematics: Applications and Interpretation and Analysis and Approaches courses, underpinning everything from portfolio risk to sampling theory.
在概率与统计中，组合随机变量是一项极为强大的建模工具。当我们研究独立随机变量时，可以方便地计算它们之和的期望与方差，并确定所得变量的分布。本主题贯穿IB文凭数学：应用与解释以及分析与方法课程，是从投资组合风险到抽样理论等众多内容的基石。
1. What Are Independent Random Variables? | 什么是独立随机变量？
all values x and y. Independence allows us to use simplified rules when adding variables.
如果随机变量X的结果不影响Y的结果，反之亦然，则称X和Y相互独立。形式上，对于所有x和y值，满足P(X=x且Y=y) = P(X=x) × P(Y=y)。独立性允许我们在进行变量加法时使用简化的运算规则。
2. Expectation of a Sum: Always Additive | 和的期望：总是可加
For any random variables X and Y, whether independent or not, the expected value of the sum is the sum of the expected values: E(X + Y) = E(X) + E(Y) e extends to linear combinations: E(aX + bY) = aE(X) + bE(Y), where a and b are constants.
对于任何随机变量X和Y，无论是否独立，和的期望值总是等于期望值之和：E(X + Y) = E(X) + E(Y)。该规则可推广到线性组合：E(aX + bY) = aE(X) + bE(Y)，其中a和b为常数。
- E(X+Y) = 3.5 + 1 = 4.5. This holds even if the die roll and coin flip are somehow linked.
- 示例：若X为公平骰子的点数（E(X)=3.5），Y为2次抛硬币中正面朝上的次数（E(Y)=1），则E(X+Y) = 3.5 + 1 = 4.5。即使骰子与硬币的投掷存在某种关联，该结果依然成立。
3. Variance of a Sum Under Independence | 独立条件下和的方差
When X and Y are independent, the variance of the sum is the sum of the variances: Var(X + Y) = Var(X) + Var(Y) = Var(X) + (-1)²Var(Y) = Var(X) + Var(Y) because adding a negative still increases spread.
当X与Y相互独立时，和的方差等于方差之和：Var(X + Y) = Var(X) + Var(Y)。这与期望不同，期望不需要独立性。对于差的情况，请注意Var(X - Y) = Var(X) + Var(-Y) = Var(X) + (-1)²Var(Y) = Var(X) + Var(Y)，因为增加负项照样会扩大离散程度。
Var(X ± Y) = Var(X) + Var(Y) (given independence)
给定独立，Var(X ± Y) = Var(X) + Var(Y)
4. General Linear Combinations: E(aX + bY) and Var(aX + bY) | 一般线性组合的期望与方差
a²Var(X) + b²Var(Y). The constant c shifts the centre but does not alter the variance. The squares a² and b² arise because variance measures squared deviations.
对于相互独立的X和Y，这些规则可推广到任何线性组合W = aX + bY + c。期望为E(W) = aE(X) + bE(Y) + c，方差为Var(W) = a²Var(X) + b²Var(Y)。常数c会平移中心，但不改变方差。出现平方项a²和b²是因为方差衡量的是平方偏差。
- If X ~ N(5, 2) and Y ~ N(3, 1), then for W = 2X - 3Y + 10: E(W)=2(5)-3(3)+10=11, Var(W)=4(2)+9(1)=17, so W ~ N(11, 17).
- 若X~N(5,2)，Y~N(3,1)，对于W=2X-3Y+10：E(W)=2(5)-3(3)+10=11，Var(W)=4(2)+9(1)=17，因此W~N(11, 17)。
5. Sum of Independent Normal Distributions | 独立正态分布的和
One of the most elegant results in probability is that the sum of independent normally distributed random variables is also normally distributed. If X ~ N(μₓ, σₓ²) and Y ~ N(μᵧ, σᵧ²) are independent, then X+Y ~ N(μₓ + μᵧ, σₓ² + σᵧ²). The normality is preserved under linear combination as well.
概率论中最优美的结论之一是，相互独立的服从正态分布的随机变量之和同样服从正态分布。若X~N(μₓ, σₓ²)且Y~N(μᵧ, σᵧ²)独立，则X+Y ~ N(μₓ + μᵧ, σₓ² + σᵧ²)。在线性组合下，正态性质依然保持。
| 变量 (Variables) | 分布 (Distribution) |
|---|---|
| X ~ N(10, 4), Y ~ N(5, 1) | X+Y ~ N(15, 5) |
| X ~ N(20, 3), Y ~ N(4, 6) | X−Y ~ N(16, 9) |
This property is essential when dealing with random errors, measurement uncertainties, and portfolio returns in finance.
当处理随机误差、测量不确定度以及金融中的投资组合收益时，该性质至关重要。
6. Sum of Independent Binomial Distributions (Same p) | 独立二项分布的和（相同p）
If X ~ B(n₁, p) and Y ~ B(n₂, p) are independent and share the same success probability p, then their sum X+Y is also binomial: X+Y ~ B(n₁ + n₂, p). This makes intuitive sense: adding the number of successes from two independent sequences of Bernoulli trials gives the total successes in n₁+n₂ trials.
若X~B(n₁, p)与Y~B(n₂, p)相互独立且具有相同的成功概率p，则它们的和也是二项分布：X+Y ~ B(n₁ + n₂, p)。这很直观：将两个独立伯努利试验序列的成功次数相加，便得到n₁+n₂次试验中的总成功次数。
E(X+Y) = (n₁+n₂)p, Var(X+Y) = (n₁+n₂)p(1-p)
E(X+Y) = (n₁+n₂)p, Var(X+Y) = (n₁+n₂)p(1-p)
If the success probabilities differ, the sum is not binomial; it follows a more complex distribution often called a Poisson-binomial. However, expectation and variance rules still apply.
若成功概率不同，其和不再服从二项分布；而是遵循一种更为复杂的分布，通常称为泊松-二项分布。然而，期望与方差的运算法则依然适用。
7. Sum of Independent Poisson Distributions | 独立泊松分布的和
Independent Poisson random variables also add nicely. If X ~ Po(λ₁) and Y ~ Po(λ₂) are independent, then X+Y ~ Po(λ₁ + λ₂). This is particularly useful in modelling arrivals: combining independent streams of events with rates λ₁ and λ₂ yields a Poisson process with rate λ₁+λ₂.
独立的泊松随机变量同样具有良好的加法性质。若X~Po(λ₁)且Y~Po(λ₂)相互独立，则X+Y ~ Po(λ₁ + λ₂)。这在到达过程建模中特别有用：将速率为λ₁和λ₂的两个独立事件流合并，将产生速率为λ₁+λ₂的泊松过程。
Example: The number of emails received per hour is X~Po(4), and the number of text messages per hour is Y~Po(3), independent. Then the total number of notifications per hour is Po(7).
示例：每小时收到的电子邮件数X~Po(4)，每小时收到的短信数Y~Po(3)，二者独立。那么每小时的通知总数为Po(7)。
8. Distribution of Sums: The Idea of Convolution | 和的分布：卷积的概念
When X and Y are independent but not from the same convenient family, we can still theoretically determine the distribution of X+Y using convolution. The probability mass function (discrete) or probability density function (continuous) of the sum is given by the convolution of the individual functions. Although IB examinations do not require manual convolution, understanding that specific distributions (Normal, Binomial, Poisson) are 'closed under addition' is vital.
当X与Y独立但不属于同一个方便的分布族时，我们依然可以通过卷积从理论上确定X+Y的分布。离散情形下的概率质量函数或连续情形下的概率密度函数由各自函数的卷积给出。尽管IB考试不要求手动计算卷积，但理解特定分布（正态、二项、泊松）对加法具有封闭性至关重要。
ately, with standard distributions we can skip integration and rely on parameter addition.
对于连续情形，概率密度函数为fₓ₊ᵧ(z) = ∫fₓ(z-y)fᵧ(y) dy。幸运的是，对于标准分布我们可以跳过积分，直接依靠参数加法。
9. Working with Sums: Problem Solving Strategies | 处理求和问题：解题策略
Typical exam questions present independent variables and ask for probabilities or parameters of their sum. The recommended steps are:
典型的考题会给出独立变量，并要求计算它们的和的概率或参数。推荐步骤如下：
- Write down the distributions of X and Y. 写下X和Y的分布。
- State the sum distribution using closed-form properties (e.g., Normal, Binomial, Poisson). 利用封闭性质声明和的分布（如正态、二项、泊松）。
- Compute E(X+Y) and Var(X+Y) directly if the distribution is not a standard one, then use normal approximation if justified (n large, p not extreme for Binomial; large λ for Poisson). 如果分布不是标准分布，则直接计算E(X+Y)和Var(X+Y)，若条件合理（二项分布n大且p非极端值；泊松分布λ大），可使用正态近似。
- Standardize to find probabilities with normal sums: Z = (S - μₛ)/σₛ. 标准化后求正态和概率：Z = (S - μₛ)/σₛ。
10. Central Limit Theorem and Sums of Many Variables | 中心极限定理与多个变量之和
The Central Limit Theorem extends the idea of adding independent variables. Regardless of the original distribution, the sum of a large number of independent and identically distributed random variables (with finite variance) is approximately normal. This is why sums and means behave so predictably in large samples.
中心极限定理拓广了独立变量相加的概念。无论原始分布如何，大量独立同分布（且方差有限）的随机变量之和近似服从正态分布。这正是大样本下和与均值行为如此可预测的原因。
², then Sₙ = ΣXᵢ has approximately a normal distribution N(nμ, nσ²), and the sample mean X̄ is approximately N(μ, σ²/n). This underlies confidence intervals and hypothesis tests.
若X₁, X₂, ..., Xₙ为独立同分布，均值为μ，方差为σ²，则总和Sₙ = ΣXᵢ近似服从正态分布N(nμ, nσ²)，样本均值X̄近似为N(μ, σ²/n)。这构成了置信区间和假设检验的理论基础。
Published by TutorHao | Mathematics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图