Bayesian Theorem Applications | IB数学：贝叶斯定理应用解析

📚 Bayesian Theorem Applications | IB数学：贝叶斯定理应用解析

The Bayes’ Theorem is one of the most powerful and practical tools in probability theory. It allows us to update our beliefs when new evidence emerges, making it essential for IB Mathematics Analysis and Approaches (AA) and Applications and Interpretation (AI) students. This article breaks down its derivation, interpretation, and exam-focused applications.

贝叶斯定理是概率论中最强大且最实用的工具之一。它使我们在获得新证据时能够更新已有信念，因此对IB数学分析与方法（AA）及应用与解释（AI）学生而言至关重要。本文将系统讲解其推导、含义以及考试导向的应用。

1. Conditional Probability Review | 条件概率回顾

Before diving into Bayes’ Theorem, we must recall the definition of conditional probability. For any two events A and B, the probability of A given B is denoted as P(A|B), and it is calculated using the formula:

在深入学习贝叶斯定理之前，我们必须回顾条件概率的定义。对于任意两个事件A和B，在B发生的条件下A发生的概率记作P(A|B)，其计算公式为：

P(A|B) = P(A ∩ B) ÷ P(B)

This formula represents the proportion of event B’s total probability that is also covered by event A. Intuitively, once we know B has occurred, the sample space shrinks to B, and we measure A’s share within that reduced space.

该公式表示事件A在事件B总概率中所占的比例。直观上，一旦我们知道B已经发生，样本空间缩小至B，我们衡量的是A在此缩减空间中所占的份额。

Additionally, the multiplication rule gives us an alternative expression: P(A ∩ B) = P(A|B) × P(B). Note that we can also write P(A ∩ B) = P(B|A) × P(A), which will be key to deriving Bayes’ Theorem.

此外，乘法法则提供了另一种表达：P(A∩B) = P(A|B) × P(B)。注意我们也可以写成P(A∩B) = P(B|A) × P(A)，这正是推导贝叶斯定理的关键所在。

2. Deriving Bayes’ Theorem | 推导贝叶斯定理

Since both expressions above represent the same intersection probability, we can set them equal to each other:

由于上面两个表达式代表相同的交集概率，我们可以令它们相等：

P(A|B) × P(B) = P(B|A) × P(A)

Dividing both sides by P(B), we obtain the fundamental form of Bayes’ Theorem:

两边同时除以P(B)，我们得到贝叶斯定理的基本形式：

P(A|B) = [P(B|A) × P(A)] ÷ P(B)

This elegant formula tells us how to “reverse” a conditional probability. If we know P(B|A) (the likelihood of evidence given a hypothesis), we can determine P(A|B) (the probability of the hypothesis given the evidence).

这个优美的公式告诉我们如何”反转”一个条件概率。如果我们知道P(B|A)（在假设成立时证据出现的似然性），就可以求出P(A|B)（在证据出现时假设成立的概率）。

3. Key Terminology | 关键术语

In exam contexts, it is crucial to understand the terminology associated with each component of Bayes’ Theorem. The terms below appear frequently in IB questions:

在考试情境中，理解贝叶斯定理各组成部分对应的术语至关重要。以下术语在IB考题中频繁出现：

Prior Probability P(A): The initial degree of belief in event A before new evidence is considered.
先验概率 P(A): 在考虑新证据之前对事件A的初始相信程度。
Likelihood P(B|A): The probability of observing evidence B assuming that A is true.
似然度 P(B|A): 假设A为真的情况下观察到证据B的概率。
Posterior Probability P(A|B): The updated probability of A after accounting for evidence B.
后验概率 P(A|B): 在考虑证据B之后A的更新概率。
Evidence P(B): The total probability of the observed evidence, often computed using the Law of Total Probability.
证据 P(B): 所观察证据的总概率，通常使用全概率公式计算。

4. The Law of Total Probability | 全概率公式

To apply Bayes’ Theorem in most real problems, we need to compute P(B), the denominator. This is where the Law of Total Probability comes in. If events A₁, A₂, …, Aₙ form a partition of the sample space (mutually exclusive and exhaustive), then:

要在大多数实际问题中应用贝叶斯定理，我们需要计算分母P(B)。这时就需要用到全概率公式。如果事件A₁, A₂, …, Aₙ构成样本空间的一个划分（互斥且完备），那么：

P(B) = P(B|A₁)P(A₁) + P(B|A₂)P(A₂) + … + P(B|Aₙ)P(Aₙ)

With this, the extended Bayes’ Theorem becomes:

由此，扩展的贝叶斯定理变为：

P(Aₖ|B) = [P(B|Aₖ)P(Aₖ)] ÷ Σᵢ P(B|Aᵢ)P(Aᵢ)

In IB exams, the partition typically involves two or three mutually exclusive hypotheses, so this sum is usually short and manageable.

在IB考试中，划分通常涉及两个或三个互斥假设，因此这个求和通常很短且易于处理。

5. Worked Example 1: Medical Testing | 例题一：医学检测

A hospital uses a test for a rare disease that affects 1% of the population. The test has a 95% sensitivity: if a person has the disease, the test is positive with probability 0.95. The test also has a 90% specificity: if a person is healthy, the test is negative with probability 0.90. If a randomly selected person tests positive, what is the probability that they actually have the disease?

某医院使用一种检测方法诊断一种影响1%人口的罕见疾病。该检测的灵敏度为95%：若一个人患病，检测呈阳性的概率为0.95。该检测的特异度为90%：若一个人健康，检测呈阴性的概率为0.90。若随机选取一个人检测呈阳性，他真正患病的概率是多少？

Let D = “has disease”, H = “healthy”, T⁺ = “positive test result”. We are given P(D) = 0.01, P(H) = 0.99, P(T⁺|D) = 0.95, and P(T⁺|H) = 0.10 (since specificity is 90%, false positive rate is 10%).

设D = “患病”，H = “健康”，T⁺ = “检测结果为阳性”。已知P(D) = 0.01，P(H) = 0.99，P(T⁺|D) = 0.95，P(T⁺|H) = 0.10（因为特异度为90%，假阳性率为10%）。

First, compute P(T⁺) using the Law of Total Probability:

首先，使用全概率公式计算P(T⁺)：

P(T⁺) = 0.95 × 0.01 + 0.10 × 0.99 = 0.0095 + 0.099 = 0.1085

Then apply Bayes’ Theorem:

然后应用贝叶斯定理：

P(D|T⁺) = (0.95 × 0.01) ÷ 0.1085 ≈ 0.0876

Surprisingly, even with a positive test result, the probability of actually having the disease is only about 8.76%. This counterintuitive result arises because the disease is rare, so most positive results are false positives.

令人惊讶的是，即使检测结果为阳性，真正患病的概率仅为约8.76%。这个违反直觉的结果源于该疾病很罕见，大多数阳性结果其实是假阳性。

6. Worked Example 2: Two Bags Problem | 例题二：双袋问题

Bag A contains 3 red and 7 blue marbles. Bag B contains 6 red and 4 blue marbles. A bag is chosen at random, and then a marble is drawn from it. The marble is red. What is the probability that it came from Bag A?

袋子A装有3颗红球和7颗蓝球。袋子B装有6颗红球和4颗蓝球。随机选择一个袋子，然后从中抽取一颗球。抽出的球是红球。问它来自袋子A的概率是多少？

Let A = “Bag A selected”, B = “Bag B selected”, R = “red marble drawn”. Since the bag is chosen at random:

设A = “选择袋子A”，B = “选择袋子B”，R = “抽到红球”。由于袋子是随机选择的：

P(A) = 0.5, P(B) = 0.5, P(R|A) = 3/10, P(R|B) = 6/10

Compute P(R):

计算P(R)：

P(R) = (3/10)(1/2) + (6/10)(1/2) = 3/20 + 6/20 = 9/20

Now apply Bayes’ Theorem:

现在应用贝叶斯定理：

P(A|R) = [(3/10)(1/2)] ÷ (9/20) = (3/20) × (20/9) = 3/9 = 1/3

There is a one-third chance the red marble came from Bag A. This makes sense because Bag A has fewer red marbles, so observing a red marble shifts our belief toward Bag B.

红球来自袋子A的概率为三分之一。这很合理，因为袋子A中红球较少，观察到红球使我们的信念向袋子B倾斜。

7. Worked Example 3: The Monty Hall Problem | 例题三：蒙提霍尔问题

In a game show, there are three doors: behind one is a car, behind the other two are goats. You pick door 1. The host, who knows what is behind each door, opens door 3 to reveal a goat, and then offers you the chance to switch to door 2. Should you switch?

在一个游戏节目中，有三扇门：一扇后面是汽车，另外两扇后面是山羊。你选择了1号门。主持人知道每扇门后面是什么，他打开了3号门，露出了一只山羊，然后给你机会换到2号门。你应该换吗？

Let C₁, C₂, C₃ denote the events that the car is behind doors 1, 2, 3 respectively. Initially, P(C₁) = P(C₂) = P(C₃) = 1/3. Let H₃ be the event that the host opens door 3.

设C₁, C₂, C₃分别表示汽车在1号、2号、3号门后面的事件。初始时P(C₁) = P(C₂) = P(C₃) = 1/3。设H₃为主持人打开3号门的事件。

We need to determine P(C₁|H₃) versus P(C₂|H₃). The key assumption is that the host always opens a door with a goat, never opens your chosen door, and if both remaining doors have goats, he chooses randomly.

我们需要确定P(C₁|H₃)与P(C₂|H₃)的大小。关键在于假设主持人总是打开一扇有山羊的门，从不打开你选的门，如果剩余两扇门都有山羊，他随机选择一扇打开。

If car is behind door 1 (your pick), then doors 2 and 3 both have goats; host opens one at random: P(H₃|C₁) = 1/2.
如果汽车在1号门后（你选的），那么2号和3号门都是山羊；主持人随机打开一扇：P(H₃|C₁) = 1/2。
If car is behind door 2, then door 3 must have a goat, so host must open door 3: P(H₃|C₂) = 1.
如果汽车在2号门后，那么3号门一定是山羊，所以主持人必然打开3号门：P(H₃|C₂) = 1。
If car is behind door 3, the host would not open door 3 because it reveals the car: P(H₃|C₃) = 0.
如果汽车在3号门后，主持人不会打开3号门，因为会暴露汽车：P(H₃|C₃) = 0。

Compute P(H₃):

计算P(H₃)：

P(H₃) = (1/2)(1/3) + (1)(1/3) + (0)(1/3) = 1/6 + 1/3 = 1/2

Then:

于是：

P(C₁|H₃) = [(1/2)(1/3)] ÷ (1/2) = 1/3

P(C₂|H₃) = [(1)(1/3)] ÷ (1/2) = 2/3

Therefore, switching doubles your chance of winning the car. This classic problem beautifully demonstrates how Bayes’ Theorem formalizes rational updating of beliefs.

因此，换门使你赢得汽车的概率翻倍。这个经典问题优美地展示了贝叶斯定理如何将理性更新信念的过程形式化。

8. Bayes’ Theorem and Tree Diagrams | 贝叶斯定理与树状图

In IB exams, students are often encouraged to use tree diagrams to visualize Bayes’ Theorem problems. A tree diagram with branches for hypotheses (first stage) and evidence (second stage) makes it easy to identify all necessary probabilities.

在IB考试中，通常鼓励学生使用树状图来可视化贝叶斯定理问题。一个包含假设分支（第一阶段）和证据分支（第二阶段）的树状图使识别所有必要的概率变得容易。

The tree diagram provides a visual representation of the Law of Total Probability: the probability of any evidence outcome is the sum of the products of probabilities along all paths leading to that outcome. Bayes’ Theorem then asks: among all paths that reach this evidence, what fraction goes through a specific hypothesis branch?

树状图提供了全概率公式的直观表示：任何证据结果出现的概率等于所有通向该结果的路径上概率乘积之和。贝叶斯定理的提问是：在所有到达该证据的路径中，有多少比例经过某个特定的假设分支？

For example, in the medical testing scenario, the tree would show D/H at the first level, and T⁺/T⁻ at the second level for each branch. The posterior probability P(D|T⁺) is simply the path D→T⁺ divided by the sum of all paths leading to T⁺.

例如，在医学检测场景中，树状图第一层显示D/H，第二层针对每个分支显示T⁺/T⁻。后验概率P(D|T⁺)就是D→T⁺路径除以所有通向T⁺的路径之和。

9. Common Pitfalls and How to Avoid Them | 常见误区及应对策略

Students frequently make several recurring mistakes when solving Bayes’ Theorem problems in IB examinations. Being aware of these pitfalls can significantly improve your accuracy.

学生在IB考试中解贝叶斯定理问题时经常犯几类重复的错误。意识到这些陷阱可以显著提高你的准确度。

Confusing P(A|B) with P(B|A): Always identify which conditional probability is given and which one is asked. They are almost never equal.
混淆P(A|B)和P(B|A)： 始终明确已知哪个条件概率、要求哪个条件概率。它们几乎从不相等。
Forgetting the prior probability: The posterior probability is a weighted combination of the likelihood and the prior. Neglecting the prior leads to incorrect results.
忘记先验概率： 后验概率是似然度和先验概率的加权组合。忽略先验会导致错误结果。
Miscomputing P(B): The denominator must include all possible ways that evidence B can occur, not just the path under consideration.
错误计算P(B)： 分母必须包含证据B发生的所有可能途径，而不只是正在考虑的路径。
Rounding too early: Keep exact fractions until the final step to avoid accumulated rounding errors.
过早四舍五入： 在最后一步之前保留精确分数，以避免累积舍入误差。

10. Applications in Real-World Contexts | 现实世界中的应用

Bayes’ Theorem is not merely an abstract mathematical concept; it has transformative applications across numerous disciplines that IB students may encounter in their studies and future careers.

贝叶斯定理不仅仅是一个抽象的数学概念；它在众多学科中有着变革性的应用，IB学生在学习和未来职业生涯中可能会遇到。

Spam Filtering: Email providers use Bayesian methods to estimate the probability that an email is spam based on the presence of certain words. P(spam|word) is computed from P(word|spam), P(word|not spam), and the base rate of spam.
垃圾邮件过滤： 电子邮件服务商使用贝叶斯方法，基于某些词语的出现来估计一封邮件是垃圾邮件的概率。P(垃圾|词语)由P(词语|垃圾)、P(词语|非垃圾)以及垃圾邮件的基础比率计算得出。
Diagnostic Testing: As shown in Example 1, Bayes’ Theorem is central to interpreting medical and engineering diagnostic tests correctly.
诊断测试： 如例1所示，贝叶斯定理解释医学和工程诊断测试结果时至关重要。
Machine Learning: Naive Bayes classifiers, which apply Bayes’ Theorem with a strong independence assumption, are widely used in text classification and sentiment analysis.
机器学习： 朴素贝叶斯分类器在文本分类和情感分析中广泛应用，它在贝叶斯定理的基础上加上强独立性假设。
Finance and Risk: Bayesian methods help analysts update the probability of default for a loan applicant as new financial information becomes available.
金融与风险： 贝叶斯方法帮助分析师在新财务信息出现时更新贷款申请人的违约概率。

11. Exam-Style Question Walkthrough | 考试风格题目精讲

Let us work through a typical IB-style exam question step by step to consolidate everything we have learned.

让我们逐步解答一道典型的IB风格考题，以巩固我们所学的一切。

Question: Factory A produces 60% of a company’s output, and Factory B produces the remaining 40%. The defect rate at Factory A is 3%, while the defect rate at Factory B is 5%. A product is selected at random and found to be defective. Find the probability that it was produced by Factory B.

题目： 工厂A生产某公司产量的60%，工厂B生产剩余40%。工厂A的次品率为3%，工厂B的次品率为5%。随机选取一件产品，发现是次品。求该产品由工厂B生产的概率。

Let A = “produced by Factory A”, B = “produced by Factory B”, D = “defective product”. We have:

设A = “由工厂A生产”，B = “由工厂B生产”，D = “次品”。已知：

P(A) = 0.60, P(B) = 0.40, P(D|A) = 0.03, P(D|B) = 0.05

Step 1: Compute P(D):

第一步：计算P(D)：

P(D) = (0.03)(0.60) + (0.05)(0.40) = 0.018 + 0.020 = 0.038

Step 2: Apply Bayes’ Theorem:

第二步：应用贝叶斯定理：

P(B|D) = [(0.05)(0.40)] ÷ 0.038 = 0.020 ÷ 0.038 ≈ 0.5263

So the probability that the defective product was produced by Factory B is approximately 52.6%.

因此次品由工厂B生产的概率约为52.6%。

Table format: Some IB questions may expect you to organize information in a table:

表格形式： 某些IB题目可能期望你用表格组织信息：

Factory	Prior P(Factory)	P(Defect\|Factory)	Joint P(Defect ∩ Factory)
A	0.60	0.03	0.018
B	0.40	0.05	0.020
Total	1.00	—	P(D) = 0.038

Then P(B|D) = 0.020 ÷ 0.038 ≈ 0.5263. Always present your working clearly to earn method marks.

然后P(B|D) = 0.020 ÷ 0.038 ≈ 0.5263。始终清晰地展示你的解题过程以获得方法分。

12. Summary of Key Formulas | 关键公式总结

To succeed in IB examinations, you must have instant recall of the following formulas. Practice them until they become second nature.

要在IB考试中取得成功，你必须能够即时回忆起以下公式。反复练习直到它们成为第二天性。

Conditional Probability:

条件概率：

P(A|B) = P(A ∩ B) ÷ P(B)

Multiplication Rule:

乘法法则：

P(A ∩ B) = P(A|B) × P(B) = P(B|A) × P(A)

Bayes’ Theorem (simple form):

贝叶斯定理（基本形式）：

P(A|B) = [P(B|A) × P(A)] ÷ P(B)

Bayes’ Theorem (extended form):

贝叶斯定理（扩展形式）：

Law of Total Probability:

全概率公式：

P(B) = Σᵢ P(B|Aᵢ)P(Aᵢ)

Mastering these five formulas, combined with careful reading of the problem statement, will allow you to handle any Bayes’ Theorem question the IB exam can throw at you.

掌握这五个公式，加上仔细审题，将使你能够应对IB考试中出现的任何贝叶斯定理题目。

Published by TutorHao | IB Mathematics Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Bayesian Theorem Applications | IB数学：贝叶斯定理应用解析