📚 3B Distributions of Functions of a Continuous Random Variable | 连续随机变量函数的分布
), where X is a continuous random variable with known probability density function (pdf) fX(x). This skill is essential for deriving sampling distributions, modelling real-world data, and performing statistical inference. Common transformations include linear scaling, squaring, exponentiating, or taking the reciprocal. In this article, we explore both the cumulative distribution function (CDF) method and the change-of-variable formula for monotonic functions, extend to non-monotonic cases, and illustrate with detailed examples relevant to IB and Cambridge A-Level Mathematics.
在概率与统计中,我们经常需要求取变换后随机变量的分布,例如 Y = g(X),其中 X 是已知概率密度函数 (pdf) fX(x) 的连续随机变量。这项技能对于推导抽样分布、建模实际数据以及进行统计推断至关重要。常见的变换包括线性缩放、平方、指数运算或取倒数。本文将探讨累积分布函数 (CDF) 方法、单调函数的变量替换公式,并将其推广到非单调情形,同时结合 IB 和剑桥 A-Level 数学相关知识点给出详细例题。
1. Why Study Distributions of Functions? | 为什么研究函数的分布?
When we transform a random variable, the new variable carries its own probability distribution. Understanding how to find this distribution enables us to compute expected values, variances, and probabilities for transformed quantities – for instance, the kinetic energy (½mv²) given a velocity distribution, or the total resistance of components connected in parallel. In statistics, pivotal quantities used in confidence intervals are functions of random variables, so mastering this topic is crucial.
当我们对随机变量进行变换时,新变量具有其自身的概率分布。掌握如何求取这种分布,使我们能够计算变换后量的期望值、方差和概率——例如,给定速度分布求动能 (½mv²),或并联元件的总电阻。在统计学中,用于置信区间的枢轴量正是随机变量的函数,因此精通本专题至关重要。
2. The Cumulative Distribution Function (CDF) Technique | 累积分布函数 (CDF) 方法
The most universal method starts from the definition of the cumulative distribution function: FY leading to an expression involving the CDF of X. Then we differentiate to obtain the pdf fY(y) = d/dy FY ormation, whether monotonic or not, as long as we can solve the inequality g(X) ≤ y for X.
最通用的方法从累积分布函数的定义出发:FY(y) = P(Y ≤ y)。对于 Y = g(X),我们将事件 {Y ≤ y} 用 X 表示,得到涉及 X 的 CDF 的表达式。然后对 y 求导得到概率密度函数 fY(y) = d/dy FY(y),前提是导数存在。该方法适用于任何变换,无论单调与否,只要能解出不等式 g(X) ≤ y 中 X 的范围即可。
Step-by-step: (1) Find the support of Y. (2) For y in the support, write FY(y) = P(g(X) ≤ y) = P(X ∈ the set of x-values satisfying g(x) ≤ y. (3) Express this probability as an integral of fX(x) over A(y). (4) Differentiate with respect to y to get fY(y). Remember to use the chain rule and Leibniz integral rule for variable limits.
具体步骤:(1) 确定 Y 的取值范围。 (2) 对取值范围内的 y,写出 FY(y) = P(g(X) ≤ y) = P(X ∈ A(y)),其中 A(y) 为满足 g(x) ≤ y 的 x 值集合。 (3) 将此概率表示为 fX(x) 在 A(y) 上的积分。 (4) 对 y 求导得到 fY(y)。注意对变上限积分使用链式法则和莱布尼茨积分法则。
3. The Change-of-Variable Formula for Monotonic Functions | 单调函数的变量变换公式
la. Let Y = g(X). The inverse function x = g⁻¹(y) exists. Then the pdf of Y is given by: fY(y) = fX(g⁻¹(y)) × |d/dy g⁻¹(y)| ises from the CDF method by differentiating, and it is often quicker when g is invertible.
若 g 在 X 的支撑集上是可导的一一映射(严格单调),则可直接使用密度变换公式。设 Y = g(X),反函数 x = g⁻¹(y) 存在。则 Y 的概率密度函数为:fY(y) = fX(g⁻¹(y)) × |d/dy g⁻¹(y)|。导数的绝对值保证了密度非负。该公式由 CDF 方法求导得出,当 g 可逆时计算更为快捷。
Published by TutorHao | IB 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