Société de Calcul Mathématique, SA
Mathematical Modelling Company Corp.

A probability book for Engineers and Researchers:

Probabilistic Information Transfer







We are glad to introduce the new book by Olga Zeydina and Bernard Beauzamy:

 

Probabilistic Information Transfer

 

A book you will be proud to have for yourself and to show to your friends !

ISBN: 978-2-9521458-6-2, ISSN : 1767-1175. Size 15,3 x 24 cm. Hardcover, 208 pages.

In real life situations, one rarely has desirably detailed information. It is sometimes incomplete, sometimes corrupted, or with missing or erroneous data. Conversely, some pieces of information do exist. Therefore, there is a natural wish: to try to use the existing information in order to reconstruct some missing items. However, this should be done with two constraints:

  • First, one should not add any artificial information, such as model assumptions (for instance, that some growth is linear, or that some law is gaussian) ;
  • Second, the result should be of probabilistic nature: we do not want a precise value for the reconstruction, but a probability law, which allows estimation of the uncertainties.
This is precisely the topic of this book. We show how to "propagate" the information, from a place where it exists to a place where we want to use it; this propagation deteriorates with the distance, somewhat as a gravitational field decreases with the distance.

The book is organized in three parts: the first part presents the basic rules, accessible with no specific expertise in probabilities; the second presents the applications to real world problems, and the third part gives the theory.

This theory comes from a situation which is rather rare these days: a new mathematical theory, entirely developed by SCM in order to meet a need which was originally expressed by the Industry, namely Framatome ANP (today Areva) in 2003. Since then, the tool was used in many cases: classifying industrial objects (Air Liquide), evaluation of pollutions (Total), estimates for water quality in European rivers (European Environment Agency), applications to the safety of nuclear reactors (Institut de Radioprotection et de Sûreté Nucléaire), and so on.

About the authors :

  • Olga Zeydina, PhD in Probabilistic Methods for Nuclear Safety, has been Research Engineer at the Société de Calcul Mathématique SA, since 2006.
  • Bernard Beauzamy, University Professor (1979-1995), founded SCM SA in 1995 and since then has been Chairman of this company.

Please read the interview in "Education Views", by Michael F. Shaughnessy, May 31, 2017:
http://www.educationviews.org/olga-zeydina-bernard-beauzamy-probabilistic-information-transfer/

This book may be bought on line, using Paypal : please click here


Table of contents

I. Introduction 9
First Part - General Presentation 15
Chapter I - Presentation of the tool 17
I. Introduction 17
II. Construction of the EPH in one dimension 18
A. Basic construction 18
B. An example of application of the EPH 21
III. Construction of the EPH in any dimension 24
A. Minimal understanding of the situation 24
B. The distance 25
C. Changes with respect to the 1-dimensional situation 25
Chapter II - Taking into account the uncertainties 33
I. Position of the problem 33
II. A simple example 34
III. Precautions to be taken 35
IV. Taking into account many uncertainties 35
Chapter III - From local to global estimates 37
I. Introduction 37
II. Investigation depending on the dimension 37
A. Low dimension 37
B. High dimension 38
III. A simple example in low dimension 40
Chapter IV - The EPH in non-homogeneous situations 45
I. Introduction 45
II. Notation 45
III. Construction of the EPH in a one-dimensional space 46
A. Before the first measurement 46
B. After the first measurement 47
C. When several measurements have been made 48
IV. A simple example 50
V. Construction of the non-homogeneous EPH in any dimension51
VI. A simple example 53
Chapter V - Putting weights upon the parameters 57
I. Introduction 57
II. Taking into account previous information 58
III. Propagating the information 58
IV. Constructing a non-isotropic EPH 59
A. The formulas 59
B. Practical construction 60
Chapter VI - How to choose the data ? 63
I. Introduction 63
II. Different pasts 65
III. Trends and randomness 66
IV. Position and speed 69
Chapter VII - One source, many sensors 73
I. Presentation 73
II. Mathematical description of the problem 73
A. Transfer function 73
B. Fusion of data 74
III. A simple example 75
IV. Other procedures 76


Second Part - Applications of the EPH 79
Chapter I - The use of the EPH in Epidemiology 81
I. Introduction 81
II. Traditional methods 81
A. Linear regression 81
B. Probability law 82
III. Description of the problem 82
IV. Building the EPH 83
A. Specific construction 83
B. Computing the parameter lambda 84
C. Fixing the parameters 85
V. Computation using EPH 85
VI. Study of increments 87
Chapter II - Applications to an Industrial Forecast 91
I. Introduction 91
II. Difference between statistics and probabilities 91
III. General comments about the use of statistics by the Industry 92
A. Regression Method 93
B. Index Point method 94
IV. Better methods 96
A. Variable Weight Method 96
B. The EPH 97
Chapter III - Estimating the pollution in a harbour 103
I. Description of the problem 103
II. Traditional methods 103
III. Using the EPH 104
A. Estimate of the pollution in the first layer 105
B. Estimates of the pollution for the second layer 107
C. Where are the risk zones ? 110
D. Deciding future measurements 112
Chapter IV - Construction of a proximity index between Industrial Objects 115
I. Description of the problem 115
II. A simple example 116
A. Non-numerical values 117
B. Normalization of the parameters 117
C. Taking into account the importance of each parameter 118
D. The distance between two objects 118
E. Construction of the proximity index 119
F. Propagation of the information using the EPH 120
Chapter V - The EPH and the safety of nuclear reactors 123
I. Description of the problem 123
II. Two dimensional problem 125
A. Step 1: transfer coefficients 126
B. Step 2: taking uncertainties into account 128
C. Step 3: introducing the EPH 130
D. Recombining the collectrons 133
E. Estimating the uncertainties 133
III. Three dimensional problem 134


Third Part - Technical Construction 135
Chapter I - General construction of the EPH 137
I. Introduction 137
II. Notation 138
III. Building the EPH in the one-dimensional case 138
A. Before the first measurement 138
B. After the first measurement 139
C. Two measurements, one dimension 154
D. Any number of measurements, one dimension 158
IV. Building the EPH in the multi-dimensional case 162
A. Bounds for each parameter 162
B. Preliminary normalization of the parameters 162
C. Densities after normalization 163
D. Case of 2 parameters. 164
E. General Case: K parameters and N measurements 165
V. Questions about the model 172
A. Discrete and continuous entropy 172
B. Dependence on the discretization 175
C. Dependence on the min and max values 176
Chapter II - Looking for dangerous zones 177
I. Introduction 177
II. Presentation of the problem 178
III. Dangerous zones 179
IV. Search for an extreme point 183
V. Random search 184
A. Preliminary warning 184
B. Uniform distribution on the unit sphere 188
C. Random search for dangerous zones 190
Chapter III - Ranking the parameters 193
I. Introduction 193
II. Practical approach 194
Chapter IV - Existing techniques 199
I. Introduction 199
II. Deterministic techniques 199
III. Probabilistic Techniques 200
A. Kriging Interpolation 200
B. Use of other information 201
References 203
Credits 205

 


In order to see the readers' comments, please click here


Order Form

I order copy/copies of the book:

"Probabilistic Information Transfer"

by Olga Zeydina and Bernard Beauzamy

Price per copy 80 Euros, including postage, shipping, handling, and VAT (when applicable, for France 5.5%).

Last name: ...

First name: ...

Company (if relevant):.

Delivery address: ...

Billing address (if different):. .

You may return this order form with your order and fax it at the number below, or you may send it by email. You may also use your own order form.

You may also use your credit card (Paypal); simply press the "Acheter" button below ("To Buy"):


Société de Calcul Mathématique SA
111 Faubourg Saint Honoré, 75008 Paris France
Phone : + 33 1 42 89 10 89,
Fax: + 33 1 42 89 10 69,
email : assistante@scmsa.eu

This book may also be bought at "Librairie Eyrolles", 61 Boulevard Saint Germain, 75005 Paris

Pour revenir à la page d'accueil, cliquer ici