IB数学AI 微积分 统计 概率 代数 高分技巧
IB Mathematics: Applications and Interpretation (AI) is designed for students who enjoy seeing how math applies to real-world problems ： from financial modeling to statistical analysis of climate data. Unlike the AA (Analysis and Approaches) course which leans toward pure theory, AI rewards practical reasoning, calculator fluency, and the ability to interpret results in context.
IB 数学应用与解释 (AI) 专为喜欢将数学应用于现实问题的学生设计：从金融建模到气候数据的统计分析。与偏向纯理论的 AA 课程不同，AI 更看重实际推理能力、计算器熟练度以及在具体情境中解读结果的能力。
1. Number and Algebra: Sequences, Series, and Financial Math 数列与金融数学
Arithmetic and geometric sequences are the bread and butter of Paper 1. Know the formulas cold: un = u1 + (n-1)d for arithmetic and un = u1 * rn-1 for geometric. The IB exam favourite is handing you a partial series and asking you to work backwards to find u1 or the common difference/ratio ： set up simultaneous equations and let your GDC do the solving.
等差数列和等比数列是 Paper 1 的基本功。牢记公式：算术数列 un = u1 + (n-1)d，等比数列 un = u1 * rn-1。IB 考试最爱给部分数列让你反推首项或公差/公比：列出方程组，用计算器求解即可。
Financial math is where AI really shines. You need to master the compound interest formula FV = PV * (1 + r/k)kn, annuity calculations for loan repayments, and the dreaded amortization tables. The key insight: every financial problem is a geometric series in disguise. The total repayment on a loan is just the sum of all individual payments discounted back to present value.
金融数学是 AI 的亮点领域。你需要掌握复利公式 FV = PV * (1 + r/k)kn、贷款还款的年金计算以及令人头疼的分期偿还表。核心思路：每个金融问题本质上都是等比数列的变形。贷款总还款额就是所有单笔付款折现到现值后求和。
SL Core Tip
Memorize the formula booklet page numbers for financial math ： the formula for TVM (Time Value of Money) is on page 6. Don't waste time deriving it in the exam.
HL Extension
HL students must handle perpetuity (n -> infinity), inflation-adjusted returns using Fisher's equation (1+r_nominal) = (1+r_real)(1+inflation), and sensitivity analysis on interest rate changes.
2. Functions and Modeling: Regression, Sinusoidal Models, and Piecewise Functions 函数建模
AI students spend a lot of time fitting models to data ： this is the course's USP (unique selling point). Your GDC is your best friend here. The regression menu (STAT -> CALC) gives you linear, quadratic, cubic, exponential, logarithmic, logistic, and sinusoidal regressions. Learn to recognize which model fits which pattern: linear for constant rate of change, exponential for proportional growth, sinusoidal for periodic data (tides, temperature, daylight hours).
AI 学生花大量时间将模型拟合到数据上：这是本课程的独特卖点。计算器是你最好的帮手。回归菜单 (STAT -> CALC) 提供线性、二次、三次、指数、对数、逻辑和正弦回归。学会识别哪种模型适合哪种模式：恒定变化率用线性，按比例增长用指数，周期数据（潮汐、温度、日照时长）用正弦。
The logistic model y = L / (1 + C*e-kx) models population growth with a carrying capacity, technology adoption, and disease spread. Key parameters: L is the upper asymptote (maximum), k controls growth rate, C shifts the curve horizontally. Always comment on the carrying capacity and inflection point when interpreting logistic regression output.
逻辑模型 y = L / (1 + C*e-kx) 可模拟有承载容量的人口增长、技术采纳和疾病传播。关键参数：L 是上渐近线（最大值），k 控制增长率，C 水平移动曲线。解读逻辑回归时务必点评承载容量和增长减速的拐点。
The Piecewise Function Trap 分段函数陷阱
Piecewise functions appear frequently in AI exams ： taxi fares, shipping costs, tax brackets. Always check the domain condition at boundary points: is the interval open (strict inequality) or closed (less-than-or-equal)? Plugging the boundary into the wrong piece costs 1-2 marks and is one of the most common IB errors. Draw a quick sketch on scratch paper if the intervals confuse you.
3. Geometry and Trigonometry: 3D, Voronoi, and Bearing Problems 几何与三角
AI's geometry section is hands-on and visual. 3D geometry involves surface area and volume calculations for prisms, pyramids, spheres, cones, and compound shapes. The angle between a line and a plane, or between two planes, uses vector dot products. Remember: the acute angle between a line and a plane is 90 degrees minus the angle between the line and the normal vector ： this twist catches many HL students off guard.
AI 的几何部分注重实操和可视化。3D 几何涉及棱柱、棱锥、球体、圆锥和复合体的表面积与体积计算。线与面的夹角或面与面的夹角使用向量点积。记住：线与面的锐角等于 90 度减去线与法向量的夹角：这个陷阱让很多 HL 学生措手不及。
Voronoi diagrams are unique to the AI syllabus. Given a set of sites (points), you partition the plane into cells where every point in a cell is closer to its site than to any other. IB questions typically ask you to: (1) find the equation of a perpendicular bisector between two sites, (2) determine which site a new point is closest to, (3) identify the "largest empty circle" within a bounded region ： this is the vertex farthest from all sites. The largest empty circle is always centered at a Voronoi vertex, never inside a cell.
Voronoi 图是 AI 课程独有的内容。给定一组站点（点），将平面划分为多个单元格，每个单元格内的任意点都更靠近其对应站点。IB 题目通常要求：(1) 求两站点间垂直平分线的方程，(2) 判断新点距离哪个站点最近，(3) 在有界区域内找出"最大空圆"：圆心必然在 Voronoi 顶点上，绝不在单元格内部。
Common Bearings Mistake
Bearings are measured clockwise from North (000 degrees to 360 degrees). The sine rule and cosine rule both apply ： but always draw a North-South-East-West diagram first. The most common error is confusing the bearing FROM A to B vs the bearing FROM B to A (they differ by 180 degrees).
HL: Adjacency and Toxcatl
HL students may face questions on Voronoi graph properties: adjacency matrix construction, the number of edges in a Voronoi diagram (Euler's formula: V - E + F = 1 for planar graphs), and applications like facility location optimization.
4. Statistics and Probability: The AI Heavyweight 统计与概率核心
This is the section where AI separates from AA ： roughly 30% of the final grade comes from statistics. Start with descriptive statistics (mean, median, mode, quartiles, IQR, standard deviation). Know how to produce a box-and-whisker plot and identify outliers using 1.5*IQR. For bivariate data: Pearson's r (linear correlation), Spearman's rank (monotonic), r2 (variance explained).统计是 AI 区别于 AA 的核心：约 30% 成绩来自统计。从描述统计入手（均值、中位数、众数、四分位数、IQR、标准差）。掌握箱线图绘制并用 1.5*IQR 判异常值。双变量数据：Pearson r（线性相关），Spearman 等级（单调相关），r2（方差解释比例）。
Probability distributions: binomial X ~ B(n, p) for counting successes; normal X ~ N(mu, sigma2) for continuous data. The crucial skill is knowing when to apply each: binomial for counting (defective items, students passing), normal for measuring (heights, weights, test scores). Poisson appears at HL for rare events (earthquakes, website visits/min).
概率分布：二项分布 X ~ B(n, p) 用于计数成功次数；正态分布 X ~ N(mu, sigma2) 用于连续数据。关键技能：计数问题用二项分布（次品数、通过考试学生数），测量问题用正态分布（身高、体重、分数）。泊松分布 HL 出现用于稀有事件（地震、网站访问量）。
Hypothesis Testing Decision Tree 假设检验决策树
SL: One-sample t-test for means; chi-squared test for independence (contingency tables) and goodness of fit. Always write H0 and H1 explicitly ： "state your hypotheses" is worth 2 marks and is the easiest 2 marks on the paper.
HL: Two-sample t-test (pooled vs unpooled variance), chi-squared for goodness of fit with estimated parameters (reduce df by 1 for each estimated parameter), ANOVA for comparing 3+ group means, and the critical distinction between Type I error (false positive, probability = alpha) and Type II error (false negative). The power of a test = 1 - P(Type II error).
5. Calculus: Differentiation, Integration, and Optimization 微积分
AI calculus is applied rather than proof-heavy. Differentiation rules (power, product, quotient, chain) are tested through real-world rates of change: velocity as the derivative of displacement, marginal cost as the derivative of total cost. The derivative f'(x) gives the instantaneous rate of change ： check units: if f(x) in meters and x in seconds, f'(x) is in m/s.
AI 微积分重应用而轻证明。求导法则（幂函数、乘法、除法、链式法则）通过实际变化率考查：速度是位移的导数，边际成本是总成本的导数。导数 f'(x) 给出瞬时变化率：检查单位：若 f(x) 以米为单位，x 以秒为单位，则 f'(x) 以 m/s 为单位。
Integration in AI focuses on definite integrals as areas under curves. The GDC handles most integration, but set up problems correctly: area between curves = integral(upper - lower), total distance = integral|velocity|, accumulated volume from flow rate. Optimization (max profit, min surface area) combines f'(x)=0 with f''(x) classification.
AI 的积分侧重于定积分作为曲线下面积。计算器处理大部分积分运算，但建立问题要正确：曲线间面积 = 积分(上-下)，总距离 = 积分|速度|，流量累积体积。优化问题（最大利润、最小表面积）结合 f'(x)=0 与 f''(x) 分类。
HL Kinematics Warning HL运动学警告
Kinematics chains differentiation and integration: s(t) -> v(t) = s'(t) -> a(t) = v'(t) = s''(t). Integrating a(t): don't forget +C ： use initial conditions! Total distance is NOT displacement when the particle changes direction. The most common HL error is skipping the constant of integration and losing the entire problem.
6. The IA: Internal Assessment Strategy 内部评估策略
The IA counts for 20% of your final IB Math AI grade ： 20 marks on mathematical communication (4), presentation (4), personal engagement (3), reflection (3), and use of mathematics (6). "Use of mathematics" carries the most weight: SL students need SL-level techniques (chi-squared, regression, t-tests); HL students must go beyond (Poisson, ANOVA, differential equations).
IA 占 IB 数学 AI 最终成绩的 20%：20 分分布在交流 (4)、表达 (4)、个人投入 (3)、反思 (3) 和数学运用 (6)。"数学运用"分值最重：SL 用 SL 技巧（卡方、回归、t 检验），HL 需超越大纲（泊松、方差分析、微分方程）。
Choose a topic with genuine personal engagement. "I play football" -> analyze shot accuracy with binomial probability. "I'm interested in climate" -> model temperature data with sinusoidal regression and hypothesis testing. Examiners have read 10,000 IAs about "height vs shoe size" ： don't be number 10,001. A focused, narrow question beats a broad, shallow one every time.
选择有真实个人投入的话题。"我踢足球" -> 用二项概率分析射门准确率。"我对气候感兴趣" -> 用正弦回归建模温度并做假设检验。考官已读一万份"身高和鞋码"的 IA：别做第一万零一份。聚焦的窄问题永远优于宽泛的浅问题。
IA Winning Formula
1. Narrow topic with personal angle
2. Collect real data (primary > secondary)
3. Apply 2-3 mathematical techniques
4. Discuss limitations honestly (reflection marks!)
5. Compare with existing literature/models
IA Red Flags to Avoid
* Pure correlation without deeper analysis
* Sample size < 30 (not statistically meaningful)
* Using only mean/mode ： too basic
* No discussion of validity/reliability
* Plagiarized background ： use your own words
7. Exam Strategy: Paper 1 vs Paper 2 vs Paper 3 (HL) 考试策略
Paper 1 (90 min SL / 120 min HL): Short-response, GDC allowed. Time management is critical ： 90 min for ~10 questions means 8-10 min each. Stuck >3 min? Circle it, move on, return later. Paper 1 is designed to be solvable: if your working gets absurdly complicated, restart from a different angle.Paper 1 (SL 90 分钟 / HL 120 分钟): 简答题，可用计算器。时间管理关键：90 分钟约 10 题，每题 8-10 分钟。卡住 >3 分钟？圈起来跳过，回头再做。Paper 1 设计来可解：如果推导异常复杂，换个角度重新开始。
Paper 2 (90 min SL / 120 min HL): Extended-response, heavier GDC use. Questions build: (a) feeds (b), feeds (c). If stuck on (a), state assumptions and proceed ： you only lose (a)'s marks. Show every GDC step: "GDC: LinReg(ax+b) L1,L2 -> a=..., b=..., r=..." not just the final equation.
Paper 2 (SL 90 分钟 / HL 120 分钟): 扩展答题，重计算器。(a) 为 (b) 铺垫，(b) 为 (c) 铺垫。若解不出 (a)，陈述假设继续 (b)：仅失 (a) 分。展示计算器步骤："GDC: LinReg(ax+b) L1,L2 -> a=..., b=..., r=..."而非仅写最终方程。
Paper 3 (HL only, 60 min): A single extended problem-solving task ： an unfamiliar scenario connecting multiple syllabus topics. Spend the first 10 minutes reading, not calculating. Identify the mathematical domains (financial math? statistics? calculus?), sketch relationships, then attack systematically. Paper 3 tests mathematical thinking, not speed.Paper 3 (仅 HL, 60 分钟): 一道扩展性问题解决任务：陌生场景连接多个大纲知识点。前 10 分钟用于阅读理解，而非计算。识别涉及的数学领域（金融数学？统计？微积分？），画出关系示意图，然后系统攻克。Paper 3 考察数学思维，而非计算速度。
8. Calculator Guide: Your GDC Is a Weapon 计算器使用指南
The fx-991CN X (or equivalent GDC) is permitted in all three papers. Master these modes: TABLE for exploring function behavior and finding roots, EQUATION for solving systems (up to 4 unknowns), STAT for all regression types and distribution calculations (normal CDF, inverse normal, binomial PDF/CDF, chi-squared CDF), MATRIX for solving transition matrices, and SPREADSHEET for organizing large datasets. A student who knows their GDC well can save 15-20 minutes across a full exam sitting compared to one who fumbles through menus.
fx-991CN X（或同级别图形计算器）在所有三张试卷中均可使用。掌握以下模式：TABLE 用于探索函数行为并找根，EQUATION 用于解方程组（最多 4 个未知数），STAT 用于所有回归类型和分布计算（正态 CDF、逆正态、二项 PDF/CDF、卡方 CDF），MATRIX 用于解转移矩阵，SPREADSHEET 用于组织大型数据集。熟练使用计算器的学生在整场考试中比在菜单中手忙脚乱的学生节省 15-20 分钟。
🛍️ 微信小店好物推荐
fx-991CN X 中文版计算器-老王推荐
¥168.00 🛒 微信小店购买9. Common Mistakes That Cost Marks 常见失分陷阱
Top 6 IB Math AI Exam Errors 六大考试错误
1Unit conversion neglect 忽略单位换算: If the problem gives data in cm and asks for volume in liters, convert BEFORE calculating. 1000 cm3 = 1 L. This costs 1-2 marks per paper ： pure carelessness.
2Incorrect GDC rounding 计算器舍入错误: Set your GDC to 3 significant figures (or 2 decimal places as required) at the start of the exam. Using full-precision intermediate values is fine, but your final answer MUST match the required precision. Writing "3.14159265359" when 3 sf is requested loses a mark.
3Confusing correlation with causation 混淆相关与因果: A high r or r2 value does not prove one variable causes the other. The IB loves docking marks for this in the IA and on Paper 2 extended responses. Always add a caveat: "this shows association, not necessarily causation ： a lurking variable may explain both."
4Forgetting the continuity correction 忘记连续性校正: When using a normal approximation to the binomial distribution, the continuity correction (+/- 0.5) is mandatory at HL. P(X >= k) becomes P(X > k-0.5) in the normal approximation. Skip this and your entire probability calculation is wrong.
5Using sample sd when you need population sd 样本标准差与总体标准差混淆: The GDC gives both sigma-x (population) and Sx (sample). In hypothesis testing with known population parameters, use sigma. For IA data (which is a sample), use Sx. Getting this wrong cascades through t-tests, confidence intervals, and chi-squared degrees of freedom.
6Incomplete hypothesis test conclusions 假设检验结论不完整: "Reject H0" earns 1 mark. Adding "at the 5% significance level" earns the second mark. Adding "there is sufficient evidence to suggest that..." in the context of the problem earns the third mark. Context is mandatory ： always link your statistical conclusion back to the real-world scenario described in the question.
10. Quick Revision Checklist 快速复习清单
Algebra 代数
Arithmetic/Geometric series formulas
Compound interest FV/PV
Loan amortization
System of 3 equations (GDC)
Functions 函数
Linear/Quad/Exp/Logistic/Sinusoidal
Regression interpretation (r, r2)
Piecewise domain conditions
Inverse function domain = range swap
Geometry 几何
3D: volume, surface area, angles
Voronoi: perpendicular bisector, largest empty circle
Bearings (N->clockwise)
Sine rule / Cosine rule / Area formula
Statistics 统计
Box plot + outlier (1.5*IQR)
Binomial B(n,p), Normal N(mu,sigma2)
t-test, chi-squared (independence + GOF)
Type I/II errors (HL)
Calculus 微积分
Power/Product/Quotient/Chain rules
Optimization: f'(x)=0 + f''(x) test
Definite integral = area
Kinematics chain (HL)
🎓 Need one-on-one tutoring? 需要一对一辅导?
📱 16621398022 同微信
Follow tutorhao on WeChat for more learning resources 关注公众号获取更多学习资源
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