Scribd
Upload
74 views2 pages

Review of Probability by Leo Breiman - S. J. Taylor

This book provides a serious introduction to modern probability theory. It avoids making the subject sterile by discussing sequences of random events before finite collections. It also does not shallow the subject by using measure theoretic machinery. The book is motivated by understanding random processes and discusses important concepts like independence, conditional probability, and martingales.

Uploaded by

Conta Do FDP
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
74 views2 pages

Review of Probability by Leo Breiman - S. J. Taylor

This book provides a serious introduction to modern probability theory. It avoids making the subject sterile by discussing sequences of random events before finite collections. It also does not shallow the subject by using measure theoretic machinery. The book is motivated by understanding random processes and discusses important concepts like independence, conditional probability, and martingales.

Uploaded by

Conta Do FDP
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

248 BOOK REVIEWS

PROBABILITY

By LEO BREIMAN: pp. ix, 421; £6.6s. (Addison-Wesley Publishing Co., London,
1968).

This book is a serious introduction to the important ideas of modern probability


theory. The author succeeds admirably in avoiding both the pitfall of writing
probability theory as a specialised branch of measure theory, which makes the subject
sterile and uninteresting; and the pitfall of avoiding all use of measure theoretic
machinery, which makes the subject shallow or non-rigorous. The book is motivated
by the desire to understand first sequences of random events, and later stochastic
processes. Thus a sequence of random variables is considered before there is any
discussion of the properties of finite collections of random variables. By this device
the reader gets quickly to the study of interesting phenomena before he has time to be
bored with the detailed development of sterile technique. However, the author
does eventually produce careful proofs of the theorems he has shown to be needed,
and the book is therefore quite satisfactory for the Mathematician interested in clear
statement and rigorous proof.
After an introductory chapter giving the motivation, the mathematical model
using (O, #", P) is introduced in Chapter 2. The important notion of independence
is studied in Chapter 3, and immediately applied to discuss zero-one laws, the law of
large numbers, and stopping times (for a sequence of random variables). There is a
carefully motivated definition of conditional probability in Chapter 4, and the
important properties of martingales are deduced in the fifth chapter with reference
to gambling systems. Chapters 6 and 7 discuss stationary processes, more general
Markov chains, and the pointwise ergodic theorem.
After a chapter mainly devoted to the tools connected with characteristic functions,
Chapter 9 introduces the central limit problem and generalises this to a discussion
of stable laws and domains of attraction. Chapter 10 is devoted to a renewal theorem
and to strengthening the central limit theorem to deal with the problem of occupation
times. After discussing discrete parameter Gaussian processes in Chapter 11, the
author goes on to define Brownian motion and obtains sample path properties and
even the strong Markov property in Chapter 12. The final four chapters discuss
the use of the in variance principle; continuous parameter martingales, and stationary
independent increment processes; and more general continuous parameter Markov
processes with particular attention to " jump " processes and diffusions.
Throughout the author assumes that his readers have a familiarity with measure
theory, so that he can use measure theoretic techniques without too much emphasis
or explanation. This allows time to devote more attention to explaining the important
probabilistic notions. The needed results from measure theory are summarised in
an appendix. Each chapter contains a supply of problems for the reader—many of
these are non-trivial—and a section of bibliographical notes. The style is chatty
throughout—as though the author were conducting a conversation. The typesetting
BOOK REVIEWS 249

is good, with few misprints observed. Regretably the price of the book is high, but
those who can bring themselves to pay six guineas for 400 pages will find this book
helpful and rewarding.
S. J. TAYLOR.

SEMI-GROUPS OF OPERATORS AND APPROXIMATION

By PAUL L. BUTZER and HUBERT BERENS: pp. xii, 318; DM.56 (Springer-Verlag,
Berlin, 1967).

This book is concerned with the theory of approximation and in particular with
the close connections between that theory and the theory of semi-groups which were
first established by the senior of the two authors, Professor Butzer.
Chapter 1 deals with the fundamentals of the theory of semi-groups; in this
context, this theory is almost entirely concerned with one-parameter semi-groups
which are continuous representations of the additive group of the non-negative
numbers. The chapter gives a very clear and concise account of this theory, including
the theory of infinitesimal generators and resolvents and the fundamental theorem
of Hille and Yosida. The theory is motivated by a discussion of initial value problems
for differential equations, particularly the heat equation, and applied to semi-groups
acting on spaces of periodic functions. Chapter 2 is concerned with the relations
between approximation theory and semi-groups. This relation stems from the fact
that most of the classical methods of approximation involve consideration of the way
in which a family of transforms of a function approximates that function: in almost
all classical cases these families form semi-groups, and the authors exploit the
connections between the behaviour of the infinitesimal generator of the semi-group
and its rate of approximation.
The classical approximation methods which converge particularly well have an
interesting property discovered by Favard and called the saturation property: they
have an optimal rate of approximation which is attained for a class of functions
called the saturation class, and converge at better than that rate only for constant
functions. This property is generalized to semi-groups.
Chapter 3 is concerned with the theory of intermediate spaces and with the inter-
polation theorems of Riesz-Thorin and Marcinkiewicz. Methods of constructing
intermediate spaces both dependent on and independent of the theory semi-groups
are given. This chapter is likely to be valuable to all students of real analysis: it is,
as far as I know, the only exposition in book form of this highly important and fast
developing theory.

You might also like