Course Summary

The course begins with the basic laws of random events and probability (sample spaces, the definition of probability, conditional probability, and independence), then introduces random variables and their numerical characteristics (expectation, variance, covariance), then uses the law of large numbers and the central limit theorem to reveal the certainty hidden behind a large number of random phenomena, and finally turns to the two inference methods of mathematical statistics: parameter estimation and hypothesis testing. Looking back over the whole course, the thread that runs through it is: from the probability of individual random events, to the overall numerical characteristics of random variables, to the limit theorems that bring certainty in the large-sample limit, and finally to estimating and testing a population from a sample.

1. The Knowledge System of the Course

TopicCore contentKey methods
Random Events and ProbabilitySample spaces and events, the definition of probability, conditional probability, independence of eventsThe addition formula, the multiplication formula, the total probability formula, Bayes’ formula
Random Variables and Their DistributionsDiscrete and continuous random variables, distribution laws and probability densities, distribution functionsDeriving a distribution law or a density, integrating to get the probability of an interval
Numerical CharacteristicsMathematical expectation, variance and standard deviation, the expectation and variance of common distributions, covariance and the correlation coefficientCompute by definition or by formula; simplify with the properties (D(X)=E(X2)−[E(X)]2D(X)=E(X^2)-[E(X)]^2)
The Law of Large Numbers and the Central Limit TheoremChebyshev’s inequality, the law of large numbers, the central limit theoremChebyshev gives an upper bound on a probability; standardise independent identically distributed variables to approximate normality
Parameter EstimationPoint estimation, the method of moments, maximum likelihood estimation, interval estimation and confidence intervalsReplace population moments with sample moments in the method of moments; differentiate the likelihood function and take its maximum
Hypothesis TestingThe idea of significance testing, the two kinds of error, tests for one and for two normal populationsPropose the hypotheses → choose a statistic → compute the critical value → draw the conclusion

2. The Law of Large Numbers and the Central Limit Theorem

The limit theorems bridge random phenomena and statistical inference, and together with the numerical characteristics they form one of the two hard-core parts of this course, so be sure to remember their conditions and conclusions.

TheoremCore resultHow to use it
Chebyshev’s InequalityP(∣X−E(X)∣≥ε)≤D(X)ε2P\big(\lvert X-E(X)\rvert\geq \varepsilon\big)\leq \dfrac{D(X)}{\varepsilon^2}An upper bound on the probability of an event follows from the expectation and the variance alone
The Law of Large NumbersUnder independent repeated trials, the sample mean Xˉ\bar{X} converges in probability to the expectation E(X)E(X)Frequencies stabilise at probabilities; it provides the theoretical basis for parameter estimation
The Central Limit TheoremThe sum of a large number of independent identically distributed random variables with finite variance, once standardised, is approximately standard normal N(0,1)N(0,1)Whatever the original distribution, the sum (or the mean) is approximately normal in the large-sample limit, which is the bedrock of statistical inference
ExpectationE(X)=∑xkpkE(X)=\sum x_k p_k (discrete) or ∫xf(x) dx\int x f(x)\,dx (continuous), E(aX+b)=aE(X)+bE(aX+b)=aE(X)+bThe expectation is a “weighted average”; linearity lets you split it up and add it up
VarianceD(X)=E(X2)−[E(X)]2D(X)=E(X^2)-[E(X)]^2, D(aX+b)=a2D(X)D(aX+b)=a^2D(X)It measures how far a variable deviates from its expectation; the variances of independent variables add
Covariance and Correlation CoefficientCov(X,Y)=E(XY)−E(X)E(Y)\mathrm{Cov}(X,Y)=E(XY)-E(X)E(Y), ρ=Cov(X,Y)D(X)D(Y)\rho=\dfrac{\mathrm{Cov}(X,Y)}{\sqrt{D(X)D(Y)}}∣ρ∣≤1\lvert\rho\rvert\leq 1; ρ=0\rho=0 means uncorrelated, and independence implies uncorrelatedness

Key takeaway: Chebyshev’s inequality “uses the variance to bound the probability”; the law of large numbers says “the sample mean settles at the expectation”; the central limit theorem says “the sum of many independent random variables is approximately normal” — the last one is the fundamental reason why the normal distribution can be used in parameter estimation and hypothesis testing.

3. How to Solve Problems

When you meet a problem in probability theory or mathematical statistics, choose a method along the following path:

  1. Draw the sample space and judge independence first: whenever conditional probability appears use the multiplication formula and the total probability formula; to “reason backwards from the result to the cause” use Bayes’ formula.
  2. Compute numerical characteristics by definition and property: for expectation and variance write the definition first, then simplify with linearity and independence; for covariance compute E(XY)E(XY) first and then subtract E(X)E(Y)E(X)E(Y).
  3. Recognise the scenario before the limit theorems: as soon as the question asks to “bound a probability from above” use Chebyshev’s inequality; when it asks for “the probability that the sum or mean of many independent identically distributed variables falls in an interval” standardise with the central limit theorem and look up the normal table.
  4. Separate point estimation from interval estimation: for point estimation use the method of moments or maximum likelihood estimation (take the logarithm of the likelihood function and then differentiate); for interval estimation choose a pivotal quantity first, then apply the confidence interval formula.
  5. Work through hypothesis testing in four steps: propose the null and alternative hypotheses → choose a suitable statistic and fix the significance level → compute the sample statistic and the critical value → if the value falls in the rejection region, reject the null hypothesis, and keep both kinds of error in mind.

Course Roadmap

  1. 1

    Exploring Functions in Advanced Mathematics

    Prerequisite

    Functions are a core idea of advanced mathematics. This course walks through foundational concepts, key properties, and classic constants so you can read, reason, and compute with confidence.

    Go to course
  2. 2

    Sequences

    Prerequisite

    Sequences bridge discrete thinking and calculus. This track covers core definitions, limits, convergence, and classic models.

    Go to course
  3. 3

    Probability and Mathematical Statistics

    Current course

    Study the laws of random phenomena, learn methods of data analysis and inference, and master the science of extracting information from data.

    Go to course

More courses

Recommended courses

Found an issue with this content?

Tell us what could be improved.