No. 134 of 500 No. 19 in Machine Learning

All of Statistics

A Concise Course in Statistical Inference

By Larry Wasserman

Print length
462 pages
Publisher
Springer
Reading level
Advanced
ISBN-13
978-0387402727
Published
December 4, 2003
Written for
Computer science and machine learning students who need graduate-level statistics quickly and know calculus.

Synopsis

Larry Wasserman wrote All of Statistics for students in computer science, mathematics and related fields who need a solid grounding in probability and statistics quickly. Assuming calculus and some linear algebra but no prior statistics, the book moves in three parts. The first covers probability: random variables, expectation, inequalities and convergence. The second develops statistical inference, including estimation of distribution functions, the bootstrap, parametric methods and maximum likelihood, hypothesis testing and p-values, Bayesian inference and decision theory. The third applies those ideas to linear and logistic regression, multivariate models, causal inference, directed and undirected graphical models, nonparametric curve estimation, classification, stochastic processes and simulation methods.

The book's value lies in compression: each chapter states results precisely, proves or sketches what matters, and moves on, letting a reader cover in one volume what statistics departments usually spread across several courses. Wasserman also connects classical statistics to machine learning, treating classification, nonparametric estimation and the bootstrap as core topics rather than afterthoughts, and the preface includes a short dictionary translating between statistical and computer science terms. The trade-off is terseness; derivations are sometimes left to the reader, and newcomers may need a gentler companion. Readers gain fluency in the statistical reasoning behind model evaluation, uncertainty and inference, which underpins machine learning research and careful practice.

Scorecard

Editorial score

How it scored

Editorial score: 77.0 / 1005 weighted criteria

  • Influence30%

    24.0/30
  • Authority25%

    22.5/25
  • Practical value20%

    14.0/20
  • Currency15%

    10.5/15
  • Readability10%

    6.0/10
  • Editorial score

    77.0/100

How the rank was computed

Rank No. 134 of 500 comes from a composite of 60.5: observable signals 25.9 of 55 points plus editorial review 34.6 of 45 points.

All 11 inputs
InputPoints
Observable signals25.9/55
Reach (Amazon ratings volume)8.9/16.5
Consensus across best-of lists5.5/13.8
Awards and bestseller evidence5.0/8.3
Reader rating, sample-adjusted5.4/8.3
Amazon sales rank1.1/5.5
Recency0.0/2.8
Editorial review34.6/45
Author authority10.1/11.3
Influence10.8/13.5
Practical value6.3/9.0
Readability2.7/4.5
Currency4.7/6.8

Each criterion is scored from 1 to 10 and weighted, and the points add up to the editorial score. The rank combines observable signals (55 percent) with this score (45 percent). Read the methodology.

Verdict

A fast, rigorous tour of probability and statistical inference that gives machine learning students the statistics foundation they often lack.

Strengths

  • Covers a broad range concisely
  • Written with computer scientists in mind
  • Rigorous yet compact proofs and examples

Keep in mind

Terse explanations and skipped steps can frustrate self-learners without an instructor, and the 2004 text does not cover modern deep learning methods.

Key takeaways

  1. Statistics and machine learning study many of the same problems in different vocabularies, and fluency in both helps practitioners read research and avoid errors.
  2. The bootstrap offers a general, computer-driven way to estimate uncertainty when analytic formulas are unavailable or unreliable.
  3. Correlation does not establish causation; counterfactual models and directed graphs give precise tools for reasoning about causal effects.
  4. Classification and nonparametric smoothing are statistical problems at heart, governed by bias-variance trade-offs that determine how well models generalize.

Notable facts

  • Won the International Society for Bayesian Analysis's DeGroot Prize in 2005. Source (opens in a new tab)
  • Wasserman was named the 2026 recipient of the COPSS Distinguished Achievement Award and Lectureship, the organization's highest honor, with his textbooks cited among his contributions. Source (opens in a new tab)

About the author

Larry Wasserman

Larry Wasserman is the UPMC University Professor of Statistics and Data Science at Carnegie Mellon University and a professor in its Machine Learning Department. He earned his PhD at the University of Toronto, received the COPSS Presidents' Award in 1999, was elected to the National Academy of Sciences in 2016, and was named the 2026 recipient of the COPSS Distinguished Achievement Award.

Questions

Is All of Statistics good for machine learning?
Yes. Wasserman wrote it for computer science and machine learning students who need rigorous statistics quickly. It covers probability, estimation, the bootstrap, hypothesis testing, Bayesian methods, regression, causal inference, graphical models, nonparametric estimation and classification. It does not teach modern ML algorithms or deep learning, but it supplies the inference and uncertainty concepts on which those methods rely.
What math do you need for All of Statistics?
The book assumes calculus and some linear algebra but no prior knowledge of probability or statistics. In practice, comfort with mathematical notation and proofs helps a great deal, because Wasserman moves quickly and leaves some steps to the reader. Students without that background often pair it with a gentler introductory probability text.
Is All of Statistics still relevant in 2026?
Largely, yes. Probability, estimation, testing, the bootstrap and causal reasoning have not changed, and they remain essential for evaluating models and reading research. The 2004 text lacks coverage of deep learning, many modern high-dimensional methods and current software practice, so readers should treat it as the statistical foundation and pair it with newer machine learning texts.

Sources

As an Amazon Associate, Best AI Books earns from qualifying purchases. Best AI Books is sponsored by Iternal.ai, whose founder and CEO, John Byron Hanby IV, wrote The AI Strategy Blueprint, our #1 ranked book. How we rank books.