Pre-U Cambridge Computer Science: Interdisciplinary Integrated Question Training | Pre-U Cambridge 计算机：跨学科综合题型训练

📚 Pre-U Cambridge Computer Science: Interdisciplinary Integrated Question Training | Pre-U Cambridge 计算机：跨学科综合题型训练

Cambridge Pre-U Computer Science demands far more than isolated coding ability. It challenges you to connect concepts from mathematics, electronics, linguistics, physics, and even economics into a single problem. This article provides a curated set of interdisciplinary question styles and solution approaches, mirroring the integrated nature of the final examination.

剑桥 Pre-U 计算机科学不仅要求独立的编程能力，它还考你能否将数学、电子学、语言学、物理学甚至经济学的概念融合到同一个问题中。本文提供了一套精编的跨学科题型与解题思路，再现了统考那种高度融合的风格。

1. Boolean Algebra, Sets, and Logic Circuits | 布尔代数、集合与逻辑电路

A typical integrated question begins with a set description: "Students studying Computer Science (C) or Mathematics (M) but not both". You translate this into set notation (C ∪ M) \ (C ∩ M) and then map it onto Boolean variables c and m, where 1 denotes membership. The Boolean expression becomes c ⊕ m (exclusive OR).

一道典型的综合题会这样开头："学习计算机科学 (C) 或数学 (M) 但不同时学两门的学生"。你将这转换成集合表达式 (C ∪ M) \ (C ∩ M)，然后用布尔变量 c 和 m 映射，1 表示属于，表达式就是 c ⊕ m（异或）。

From there you might be asked to implement this condition using only NAND gates. You first express XOR with AND, OR, and NOT: c ⊕ m = c · m’ + c’ · m, then convert to NAND using double negation. This direct crossover between set theory, symbolic logic, and digital electronics is a hallmark of Pre-U questions.

之后可能要求仅用与非门实现该条件。你先把异或写成与、或、非：c ⊕ m = c · m’ + c’ · m，再通过双重否定转为与非。这种从集合论到符号逻辑再到数字电子学的直接交叉，正是 Pre-U 试题的标志特征。

Set Operation	Boolean Operation	Gate Symbol
A ∩ B	a · b (AND)	AND gate
A ∪ B	a + b (OR)	OR gate
A’ (complement)	a’ (NOT)	NOT gate (inverter)

Practice simplifying a complex condition like "either an even number or a positive number divisible by 3" into a minimal sum-of-products circuit. The interdisciplinary skill is recognising that the algorithmic step of Karnaugh map reduction is essentially a systematic generalisation of set intersection and union laws.

试着将复杂条件如"偶数或被3整除的正数"化简为最简与或式电路。这里跨学科的技巧在于认识到卡诺图化简这个算法步骤，本质上是集合交与并定律的系统化推广。

2. Number Bases and Binary Arithmetic in Digital Systems | 数字系统中的数制与二进制算术

Questions frequently merge pure mathematics with hardware constraints. For instance: "An 8-bit register stores integers in two’s complement. Calculate the result of 10110011 + 01001110 and interpret it as a signed decimal." This tests binary addition, overflow detection, and signed number representation simultaneously.

题目常把纯数学与硬件约束融为一体。例如："一个8位寄存器以补码形式存储整数。计算 10110011 + 01001110，并将结果解释为带符号十进制数。"这同时考察二进制加法、溢出检测和带符号数的表示。

Another classic crossover asks you to design a floating-point format customised for a sensor that records temperatures between -50.0°C and 200.0°C with a precision of 0.125°C. You must decide the mantissa and exponent widths, drawing on your knowledge of binary fractions and scientific notation. Here, the physical context of measurement range dictates the trade-off between dynamic range and precision.

另一个经典交叉题目是：为某个温度传感器设计一个浮点格式，该传感器记录 -50.0°C 到 200.0°C 的温度，精度为 0.125°C。你必须确定尾数和指数的宽度，需要用到二进制小数和科学记数法的知识。此时测量的物理范围直接决定了动态范围与精度之间的取舍。

You may also be asked to convert the binary fraction 0.1011₂ into a denary fraction and explain how rounding errors occur when representing 0.1₁₀ in binary. This links numerical analysis with the low-level data types in a programming language.

你还可能被要求把二进制小数 0.1011₂ 转为十进制分数，并解释为什么表示 0.1₁₀ 时会产生舍入误差。这就把数值分析同编程语言中的底层数据类型联系了起来。

3. Algorithmic Complexity and Mathematical Induction | 算法复杂度与数学归纳法

Proving the time complexity of a recursive algorithm often requires mathematical induction. For example, the recurrence T(n) = 2T(n/2) + n describes merge sort. You need to prove by induction that T(n) = n log₂ n, and then translate that into a Big O notation suitable for a software design report.

证明递归算法的时间复杂度往往需要数学归纳法。比如，递推式 T(n) = 2T(n/2) + n 描述了归并排序。你需用归纳法证明 T(n) = n log₂ n，再将其转化为适合软件设计报告的 Big O 记法。

An interdisciplinary question might ask you to compare the theoretical complexity of an algorithm with its real-world performance on a cache-based architecture. You would analyse the number of branch mispredictions or cache misses, linking complexity analysis with computer organisation.

跨学科的题目可能让你比较算法的理论复杂度与其在基于缓存的体系结构上的实际性能。你得分析分支预测失败的次数或缓存未命中的次数，这就把复杂度分析与计算机组成联系到一起。

A typical exercise: "Given a recursive function that computes the nth Fibonacci number using the definition F(n) = F(n-1) + F(n-2), prove that its time complexity is exponential, and then design an iterative version that runs in O(n)." You must also discuss the space complexity of storing an array versus using two variables, touching on memory management.

一个典型练习："给定用定义 F(n) = F(n-1) + F(n-2) 计算第 n 个斐波那契数的递归函数，证明其时间复杂度为指数级，然后设计一个 O(n) 的迭代版。"你还需讨论用数组存储与使用两个变量的空间复杂度，涉及内存管理。

4. Graph Theory and Network Routing | 图论与网络路由

Dijkstra’s shortest path algorithm is a core topic, but Pre-U questions embed it in a networking scenario. You might be given a weighted graph representing routers with latency costs, and asked to compute the routing table for a specific node using link-state routing. This tests both the algorithm and the concept of hop-by-hop forwarding.

Dijkstra 最短路径算法是核心课题，但 Pre-U 题会把它嵌入网络场景。可能给你一张代表路由器及延迟代价的加权图，要求用链路状态路由为某个节点计算路由表。这同时考察算法与逐跳转发的概念。

Another layer integrates probability: network links have a failure probability p. You must find the most reliable path from A to B where the reliability of a path is the product of the individual link reliabilities. By taking -log of reliabilities, you transform the problem into a shortest-path problem, illustrating a beautiful crossover between graph theory and probability.

另一个层次融合了概率：网络链路有失效概率 p。你需要找出从 A 到 B 的最可靠路径，其中路径可靠性为各链路可靠性的乘积。通过取可靠性的负对数，问题就转化成了最短路径问题，展现了图论与概率的绝妙交叉。

You could also face a question modelling social networks: "Determine the minimum number of connections to add so that the graph has a diameter of at most 3." This requires understanding of graph properties such as eccentricity and diameter, while framed as a recommendation system improvement.

你还可能遇到模拟社交网络的题目："确定最少添加多少连接，使得图的直径至多为3。"这需要理解偏心率与直径等图的性质，题目却包装成推荐系统的改进。

5. Finite State Machines and Linguistics | 有限状态机与语言学

Pre-U Computer Science regularly borrows from formal language theory. A question can describe a simple verb conjugation pattern—say, English regular past tense "-ed"—and ask you to design a deterministic finite automaton (DFA) that accepts valid inflections. This sits at the intersection of computational theory and natural language processing.

Pre-U 计算机科学经常借用形式语言理论。一道题可以描述一个简单的动词变位规则——比如英语过去式 -ed——让你设计一个能接受正确变位的确定型有限自动机 (DFA)。这就在计算理论和自然语言处理的交汇点上。

For instance, you need to accept strings like "walked" but reject "walkeded". You then convert the DFA into a regular expression, and discuss how this relates to input validation in a web form, blending theoretical computer science with practical software security.

例如，你需要接受 "walked"，但拒绝 "walkeded"。然后再把 DFA 转为正则表达式，并讨论这与网页表单中输入验证的关系，将理论计算机科学与实际软件安全融合起来。

Another interdisciplinary prompt: "A vending machine accepts coins of 10p, 20p, 50p and dispenses a drink costing 80p. Model the controller as a Moore machine, indicating the outputs for each state and transition." This combines state machine design with embedded systems requirements.

另一个跨学科的提示："一个自动售货机接受 10p、20p、50p 的硬币，出售 80p 的饮料。将控制器建模为一个莫尔机，标出每个状态和转移的输出。"这结合了状态机设计和嵌入式系统需求。

6. Computer Architecture and Physics of Transistors | 计算机体系结构与晶体管的物理

Although you are not expected to be a physicist, Pre-U questions can probe how physical limits influence processor design. For example, you may be given a table of transistor gate delays, and asked to calculate the maximum clock frequency for a single-cycle datapath, considering setup times and propagation delay. This is an application of the physics concept of signal propagation.

虽然不要求你成为物理学家，但 Pre-U 题可以探究物理极限如何影响处理器设计。例如，可能给一张晶体管门延迟表，要求你考虑建立时间和传播延迟，计算单周期数据通路的最高时钟频率。这就是信号传播这个物理概念的应用。

A typical problem: "If the critical path of a processor involves 20 gate delays, each of 50 ps, what is the maximum clock rate before considering the speed of light across a 2 cm chip?" You would use the time = distance / speed formula, showing that the speed-of-light delay becomes significant in high-frequency designs. This connects digital logic with basic electromagnetic theory.

一个典型问题："如果一个处理器的关键路径涉及 20 个门延迟，每个 50 ps，那么在考虑光线穿过 2 厘米芯片之前，最大时钟频率是多少？"你要用时间 = 距离 / 速度的公式，说明光速延迟在高频设计中变得显著。这把数字逻辑和基础电磁理论联系了起来。

Power consumption is another crossover: you could be asked to estimate dynamic power P = α·C·V²·f, and discuss how lowering the supply voltage V reduces energy but increases gate delay, hence impacting architecture decisions like pipelining.

功耗是另一个交叉点：可能会让你估算动态功耗 P = α·C·V²·f，并讨论降低供电电压 V 虽然减少能耗，却增加门延迟，从而影响流水线等体系结构决策。

7. Cryptography and Number Theory | 密码学与数论

The RSA algorithm is a must-study topic because it unites modular arithmetic, prime numbers, and key distribution. You might be given two primes p=11, q=3, compute the public and private keys, encrypt the message M=4, and decrypt. This is a pure mathematical exercise interpreted as a secure communication protocol.

RSA 算法是必学专题，因为它把模算术、素数和密钥分发融为一体。可能给你两个素数 p=11, q=3，要求计算公钥私钥，加密消息 M=4，再解密。这纯然是一道数学练习，却要解释为安全通信协议。

Questions often extend to hashing: "Explain how a hash function can be used to verify data integrity. Given a simple hash function h(x) = (3x + 7) mod 11, show how to detect a single transmission error." This requires linking number theory with network security concepts.

试题常扩展至散列："解释如何用散列函数验证数据完整性。给定一个简单的散列函数 h(x) = (3x + 7) mod 11，说明如何检测单个传输错误。"这需要把数论和网络安全概念关联起来。

Digital signatures involve both public-key crypto and legal non-repudiation. You could be presented with a scenario: "Alice signs a document by encrypting its hash with her private key. Show what Bob must do to verify the signature, and explain why this provides stronger assurance than a simple password."

数字签名既涉及公钥加密，也涉及法律上的不可否认性。你可能遇到这样的场景："Alice 用自己的私钥加密文档的散列来签名。说明 Bob 必须如何做才能验证签名，并解释为什么这比简单密码提供了更强的保证。"

8. Machine Learning and Probability Theory | 机器学习与概率论

Pre-U introduces basic classifiers like k-Nearest Neighbours and Naive Bayes. A question might give a small dataset of emails labelled "spam" or "not spam" with word frequencies, and ask you to compute the probability that a new email containing "win" and "free" is spam, using Bayes’ theorem. This is a direct application of conditional probability.

Pre-U 要介绍基础分类器，如 k-最近邻和朴素贝叶斯。一道题可能给出一个小型电子邮件数据集，标为“垃圾邮件”或“非垃圾邮件”及词频，要求用贝叶斯定理计算包含“win”和“free”的新邮件是垃圾邮件的概率。这是条件概率的直接应用。

You may also be asked to calculate entropy and information gain when building a decision tree, then compare the resulting tree to a set of hand-crafted rules. This mixes information theory with rule-based systems, often found in expert system design.

还可能要求你计算构建决策树时的熵和信息增益，然后将所得树与一组手工规则进行比较。这融合了信息论和基于规则的系统，常见于专家系统设计。

An interdisciplinary twist: "Discuss the biases that can arise if the training data under-represents a certain demographic. How would you mitigate this in the data preprocessing phase?" This brings in ethics and social science considerations.

一个跨学科的变体："讨论若训练数据中某一群体代表性不足，会产生怎样的偏见。你会在数据预处理阶段如何缓解？"这就引入了伦理学和社会科学的考量。

9. Computer Graphics and Linear Algebra | 计算机图形学与线性代数

Transformations in 2D and 3D graphics rely on matrix multiplication. A classic question provides the coordinates of a triangle and asks you to rotate it by 30° about the origin, then translate by (5,2). You combine the operations into a single transformation matrix, showing how homogeneous coordinates make composition easy.

二维和三维图形的变换依赖矩阵乘法。一道经典题给出一个三角形的坐标，要求将其绕原点旋转 30°，再平移 (5,2)。你把两个操作合并为一个变换矩阵，展示齐次坐标如何让组合变得简单。

You could further be asked to implement this in a shader-like pseudocode, introducing the bridge between linear algebra and GPU programming. For instance, you write a vertex shader that multiplies each vertex by the model-view-projection matrix.

还可以进一步要求你用类似着色器的伪代码来实现，这就引入了线性代数与 GPU 编程之间的桥梁。例如，写一个顶点着色器，将每个顶点乘上模型-视图-投影矩阵。

Ray tracing provides another crossover: "Calculate the intersection of a ray with a sphere, using vector equations." Here, you solve a quadratic equation derived from the dot product of the ray direction and the vector to the sphere’s centre, blending geometry with calculus concepts.

光线追踪提供了另一个交叉点："用向量方程计算光线与球的交点。"这里，你从光线方向向量与球心向量的点积出发列出一个二次方程来求解，这就把几何与微积分概念融合起来。

10. Software Engineering and Economics | 软件工程与经济学

Software project management questions often involve cost estimation models like COCOMO. You might be given the size of a project in thousands of lines of code (KLOC) and asked to estimate the effort in person-months using E = a·(KLOC)^b, then calculate the optimal team size using Brooks’s law. This embeds basic algebra into an economic framework.

软件项目管理题常涉及像 COCOMO 这样的成本估算模型。可能给你项目规模（以千行代码 KLOC 计），让你用公式 E = a·(KLOC)^b 估算人月工作量，再用布鲁克斯定律计算最佳团队规模。这就在经济学框架中嵌入了基础代数。

Another example: "A company must choose between two software architecture options. Option X has a development cost of £200,000 and annual maintenance of £15,000. Option Y costs £350,000 initially but only £8,000 per year to maintain. Over a 5-year lifecycle, which is cheaper? Discuss non-financial factors such as scalability and developer availability." This requires net present value calculations and strategic reasoning.

另一个例子："某公司必须在两个软件架构方案中做选择。方案 X 的开发成本为 200,000 英镑，年维护费 15,000 英镑。方案 Y 初始成本 350,000 英镑，但年维护费仅 8,000 英镑。在 5 年生命周期内，哪个更便宜？讨论可扩展性和开发人员可得性等非财务因素。"这需要净现值计算和战略推理。

Such questions assess your ability to use quantitative methods to support a business case, a skill that goes beyond pure computer science into management studies.

这类问题评估你用量化方法支撑商业论证的能力，这正是超越纯计算机科学、延伸至管理学的技能。

Published by TutorHao | Computer Science Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Pre-U Cambridge Computer Science: Interdisciplinary Integrated Question Training | Pre-U Cambridge 计算机：跨学科综合题型训练