Namespaces
Variants
Actions

Difference between revisions of "Bonferroni-type inequalities"

From Encyclopedia of Mathematics
Jump to: navigation, search
(Importing text file)
 
m (tex done)
 
Line 1: Line 1:
The [[Bonferroni inequalities|Bonferroni inequalities]] may become impractical as a result of the large number of terms in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b1107401.png" /> (for notation, see [[Bonferroni inequalities|Bonferroni inequalities]]). In order to overcome this difficulty, two kinds of modification can be performed:
+
{{TEX|done}}
  
i) one can limit the number of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b1107402.png" /> used in a bound but combine this with "better" coefficients than would be provided by a Bonferroni bound;
+
The [[Bonferroni inequalities|Bonferroni inequalities]] may become impractical as a result of the large number of terms in  $ S _ {k} $(
 +
for notation, see [[Bonferroni inequalities|Bonferroni inequalities]]). In order to overcome this difficulty, two kinds of modification can be performed:
  
ii) one can modify the definition of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b1107403.png" /> by limiting its number of terms, but only to an extent that still produces the Bonferroni bounds for some <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b1107404.png" />. The new inequalities obtained in this manner are called Bonferroni-type inequalities. In other words, for i), given the values <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b1107405.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b1107406.png" />, where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b1107407.png" /> is a given finite set, one seeks coefficients <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b1107408.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b1107409.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074010.png" />, and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074011.png" /> such that, whatever the [[Probability space|probability space]] and the events <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074012.png" /> on it,
+
i) one can limit the number of $  S _ {k} $
 +
used in a bound but combine this with  "better" coefficients than would be provided by a Bonferroni bound;
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074013.png" /></td> <td valign="top" style="width:5%;text-align:right;">(a1)</td></tr></table>
+
ii) one can modify the definition of  $  S _ {k} $
 +
by limiting its number of terms, but only to an extent that still produces the Bonferroni bounds for some  $  0 \leq  r \leq  n $.  
 +
The new inequalities obtained in this manner are called Bonferroni-type inequalities. In other words, for i), given the values  $  S _ {k} $,
 +
$  k \in T $,
 +
where  $  T $
 +
is a given finite set, one seeks coefficients  $  a _ {k} ( n,r ) $,
 +
b _ {k} ( n,r ) $,
 +
$  c _ {k} ( n,r ) $,
 +
and  $  d _ {k} ( n,r ) $
 +
such that, whatever the [[Probability space|probability space]] and the events  $  A _ {j} $
 +
on it,
 +
 
 +
$$ \tag{a1}
 +
\sum _ { T } a _ {k} ( n,r ) S _ {k} \leq  {\mathsf P} ( m _ {n} ( A ) \geq  r ) \leq  \sum _ { T } b _ {k} ( n,r ) S _ {k} ,
 +
$$
  
 
and
 
and
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074014.png" /></td> <td valign="top" style="width:5%;text-align:right;">(a2)</td></tr></table>
+
$$ \tag{a2}
 +
\sum _ { T } c _ {k} ( n,r ) S _ {k} \leq  {\mathsf P} ( m _ {n} ( A ) = r ) \leq  \sum _ { T } d _ {k} ( n,r ) S _ {k} ,
 +
$$
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074015.png" /> signifies summation over <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074016.png" />. In the following simple example, all problems and difficulties arising in the solution of (a1) and (a2) are demonstrated.
+
where $  \sum _ {T} $
 +
signifies summation over $  k \in T $.  
 +
In the following simple example, all problems and difficulties arising in the solution of (a1) and (a2) are demonstrated.
  
Let <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074017.png" />, so that one can use <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074018.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074019.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074020.png" /> in setting the bounds in (a1) or (a2). Concentrating on (a2) and considering only <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074021.png" />, one looks for coefficients <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074022.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074023.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074024.png" />, and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074025.png" /> (it easily follows that <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074026.png" />) such that (a2) be valid. It turns out (and it is far from easy to arrive at such a conclusion; see [[#References|[a7]]] and [[#References|[a2]]]) that with this choice of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074027.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074028.png" /> no better bounds can be found than
+
Let $  T = \{ 0,1,2 \} $,
 +
so that one can use $  S _ {0} = 1 $,  
 +
$  S _ {1} $
 +
and $  S _ {2} $
 +
in setting the bounds in (a1) or (a2). Concentrating on (a2) and considering only $  r = 0 $,  
 +
one looks for coefficients $  c _ {1} ( n,0 ) $,
 +
$  c _ {2} ( n,0 ) $,
 +
$  d _ {1} ( n,0 ) $,  
 +
and $  d _ {2} ( n,0 ) $(
 +
it easily follows that $  c _ {0} ( n,0 ) = d _ {0} ( n,0 ) = 1 $)  
 +
such that (a2) be valid. It turns out (and it is far from easy to arrive at such a conclusion; see [[#References|[a7]]] and [[#References|[a2]]]) that with this choice of $  T $
 +
and $  r $
 +
no better bounds can be found than
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074029.png" /></td> </tr></table>
+
$$
 +
1 - S _ {1} + {
 +
\frac{2}{n}
 +
} S _ {2} \leq  {\mathsf P} ( m _ {n} ( A ) = 0 ) \leq
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074030.png" /></td> </tr></table>
+
$$
 +
\leq 
 +
1 - {
 +
\frac{2}{j + 1}
 +
} S _ {1} + {
 +
\frac{2}{j ( j + 1 )}
 +
} S _ {2} ,
 +
$$
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074031.png" /> is an arbitrary integer.
+
where $  1 \leq  j \leq  n - 1 $
 +
is an arbitrary integer.
  
For (a2), the following remarkable result has been established in [[#References|[a4]]]: It is valid on an arbitrary probability space for every choice of the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074032.png" /> if and only if it is valid for independent <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074033.png" /> with each <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074034.png" /> either equal to some <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074035.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074036.png" />, or zero. On the other hand, it was proved in [[#References|[a6]]] that if (a1) is valid for the independent structures just mentioned, then (a1) is valid on an arbitrary probability space whatever the choice of the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074037.png" />. The above results reduce both (a1) and (a2) to polynomial inequalities. Amongst the several advantages of transforming (a1) and (a2) to polynomials, known as the method of polynomials for proving Bonferroni-type inequalities, is the following reduction formula: Assume one has found coefficients <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074038.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074039.png" /> such that (a2) holds for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074040.png" /> and some set <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074041.png" />. Then both (a1) and (a2) hold for arbitrary <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074042.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074043.png" />, with the set <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074044.png" /> obtained from <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074045.png" /> by adding <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074046.png" /> to each of its elements, and the coefficients of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074047.png" /> in (a1) and (a2) satisfy the relations
+
For (a2), the following remarkable result has been established in [[#References|[a4]]]: It is valid on an arbitrary probability space for every choice of the $  A _ {j} $
 +
if and only if it is valid for independent $  A _ {j} $
 +
with each $  {\mathsf P} ( A _ {j} ) $
 +
either equal to some $  p $,  
 +
0 \leq  p \leq  1 $,  
 +
or zero. On the other hand, it was proved in [[#References|[a6]]] that if (a1) is valid for the independent structures just mentioned, then (a1) is valid on an arbitrary probability space whatever the choice of the $  A _ {j} $.  
 +
The above results reduce both (a1) and (a2) to polynomial inequalities. Amongst the several advantages of transforming (a1) and (a2) to polynomials, known as the method of polynomials for proving Bonferroni-type inequalities, is the following reduction formula: Assume one has found coefficients $  c _ {k} ( n ) = c _ {k} ( n,0 ) $
 +
and $  d _ {k} ( n ) = d _ {k} ( n,0 ) $
 +
such that (a2) holds for $  r = 0 $
 +
and some set $  T = T _ {0} $.  
 +
Then both (a1) and (a2) hold for arbitrary $  r $,  
 +
0 \leq  r \leq  n $,  
 +
with the set $  T = T _ {r} $
 +
obtained from $  T _ {0} $
 +
by adding $  r $
 +
to each of its elements, and the coefficients of $  S _ {k + r,n + r}  $
 +
in (a1) and (a2) satisfy the relations
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074048.png" /></td> </tr></table>
+
$$
 +
c _ {k + r}  ( n + r,r ) = \left ( \begin{array}{c}
 +
{k + r} \\
 +
r
 +
\end{array}
 +
\right ) c _ {k} ( n ) ,
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074049.png" /></td> </tr></table>
+
$$
 +
d _ {k + r}  ( n + r,r ) = \left ( \begin{array}{c}
 +
{k + r} \\
 +
r
 +
\end{array}
 +
\right ) d _ {k} ( n ) .
 +
$$
  
Upon replacing <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074050.png" /> by <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074051.png" /> in the right-hand sides above, one obtains <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074052.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074053.png" />, respectively.
+
Upon replacing $  \left ( \begin{array}{c}
 +
{k + r} \\
 +
r
 +
\end{array}
 +
\right ) $
 +
by $  \left ( \begin{array}{c}
 +
{k + r - 1} \\
 +
{r - 1}
 +
\end{array}
 +
\right ) $
 +
in the right-hand sides above, one obtains $  a _ {k + r}  ( n + r,r ) $
 +
and b _ {k+r}  ( n + r,r ) $,  
 +
respectively.
  
For modification ii) of the Bonferroni inequalities there are very general and very effective graph sieves. Choose an arbitrary [[Graph|graph]] with vertices <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074054.png" /> (that is, the edge set is chosen in an arbitrary manner). Now define <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074055.png" /> by deleting many terms from <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074056.png" />; in fact, retain in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074057.png" /> only those terms of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074058.png" /> for which <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074059.png" /> has no edge from the underlying graph just chosen if <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074060.png" /> is even, and allow, for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074061.png" /> odd, a well-defined number of edges in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074062.png" />, the number of which may depend on <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074063.png" /> but not on <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074064.png" />. Define <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074065.png" /> in a similar manner, but interchange the role of  "even"  and  "odd" . It is proved in [[#References|[a8]]], for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074066.png" />, and in [[#References|[a1]]], for arbitrary <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074067.png" />, that the Bonferroni lower bounds remain valid with <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074068.png" />, whilst the upper bounds remain valid with <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074069.png" />. These new bounds found many applications in random set selections, information theory, and extreme value theory; see [[#References|[a5]]]. The basic idea of constructing <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074070.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b110/b110740/b11074071.png" /> is similar to Brun's method in number theory (cf. [[Brun sieve|Brun sieve]]).
+
For modification ii) of the Bonferroni inequalities there are very general and very effective graph sieves. Choose an arbitrary [[Graph|graph]] with vertices $  \{ 1 \dots n \} $(
 +
that is, the edge set is chosen in an arbitrary manner). Now define $  S _ {k}  ^ {*} $
 +
by deleting many terms from $  S _ {k} $;  
 +
in fact, retain in $  S _ {k + r}  ^ {*} $
 +
only those terms of $  S _ {k + r}  $
 +
for which $  ( i _ {1} \dots i _ {k + r}  ) $
 +
has no edge from the underlying graph just chosen if $  k $
 +
is even, and allow, for $  k $
 +
odd, a well-defined number of edges in $  ( i _ {1} \dots i _ {k + r}  ) $,  
 +
the number of which may depend on $  r $
 +
but not on $  k $.  
 +
Define $  S _ {k} ^ {* *} $
 +
in a similar manner, but interchange the role of  "even"  and  "odd" . It is proved in [[#References|[a8]]], for $  r = 0 $,  
 +
and in [[#References|[a1]]], for arbitrary $  r $,  
 +
that the Bonferroni lower bounds remain valid with $  S _ {k}  ^ {*} $,  
 +
whilst the upper bounds remain valid with $  S _ {k} ^ {* *} $.  
 +
These new bounds found many applications in random set selections, information theory, and extreme value theory; see [[#References|[a5]]]. The basic idea of constructing $  S _ {k}  ^ {*} $
 +
and $  S _ {k} ^ {* *} $
 +
is similar to Brun's method in number theory (cf. [[Brun sieve|Brun sieve]]).
  
 
Extension of Bonferroni and Bonferroni-type inequalities have also been studied in multivariate settings. Here, multivariate means that one faces several sequences of events, and one establishes linear bounds on the joint distribution of the numbers counting the occurrences in each sequence of events. Bounds are again in terms of binomial moments. Such multivariate studies are quite new; see [[#References|[a3]]]. For a full coverage of the known (1996) multivariate bounds, see [[#References|[a5]]].
 
Extension of Bonferroni and Bonferroni-type inequalities have also been studied in multivariate settings. Here, multivariate means that one faces several sequences of events, and one establishes linear bounds on the joint distribution of the numbers counting the occurrences in each sequence of events. Bounds are again in terms of binomial moments. Such multivariate studies are quite new; see [[#References|[a3]]]. For a full coverage of the known (1996) multivariate bounds, see [[#References|[a5]]].

Latest revision as of 12:05, 9 February 2020


The Bonferroni inequalities may become impractical as a result of the large number of terms in $ S _ {k} $( for notation, see Bonferroni inequalities). In order to overcome this difficulty, two kinds of modification can be performed:

i) one can limit the number of $ S _ {k} $ used in a bound but combine this with "better" coefficients than would be provided by a Bonferroni bound;

ii) one can modify the definition of $ S _ {k} $ by limiting its number of terms, but only to an extent that still produces the Bonferroni bounds for some $ 0 \leq r \leq n $. The new inequalities obtained in this manner are called Bonferroni-type inequalities. In other words, for i), given the values $ S _ {k} $, $ k \in T $, where $ T $ is a given finite set, one seeks coefficients $ a _ {k} ( n,r ) $, $ b _ {k} ( n,r ) $, $ c _ {k} ( n,r ) $, and $ d _ {k} ( n,r ) $ such that, whatever the probability space and the events $ A _ {j} $ on it,

$$ \tag{a1} \sum _ { T } a _ {k} ( n,r ) S _ {k} \leq {\mathsf P} ( m _ {n} ( A ) \geq r ) \leq \sum _ { T } b _ {k} ( n,r ) S _ {k} , $$

and

$$ \tag{a2} \sum _ { T } c _ {k} ( n,r ) S _ {k} \leq {\mathsf P} ( m _ {n} ( A ) = r ) \leq \sum _ { T } d _ {k} ( n,r ) S _ {k} , $$

where $ \sum _ {T} $ signifies summation over $ k \in T $. In the following simple example, all problems and difficulties arising in the solution of (a1) and (a2) are demonstrated.

Let $ T = \{ 0,1,2 \} $, so that one can use $ S _ {0} = 1 $, $ S _ {1} $ and $ S _ {2} $ in setting the bounds in (a1) or (a2). Concentrating on (a2) and considering only $ r = 0 $, one looks for coefficients $ c _ {1} ( n,0 ) $, $ c _ {2} ( n,0 ) $, $ d _ {1} ( n,0 ) $, and $ d _ {2} ( n,0 ) $( it easily follows that $ c _ {0} ( n,0 ) = d _ {0} ( n,0 ) = 1 $) such that (a2) be valid. It turns out (and it is far from easy to arrive at such a conclusion; see [a7] and [a2]) that with this choice of $ T $ and $ r $ no better bounds can be found than

$$ 1 - S _ {1} + { \frac{2}{n} } S _ {2} \leq {\mathsf P} ( m _ {n} ( A ) = 0 ) \leq $$

$$ \leq 1 - { \frac{2}{j + 1} } S _ {1} + { \frac{2}{j ( j + 1 )} } S _ {2} , $$

where $ 1 \leq j \leq n - 1 $ is an arbitrary integer.

For (a2), the following remarkable result has been established in [a4]: It is valid on an arbitrary probability space for every choice of the $ A _ {j} $ if and only if it is valid for independent $ A _ {j} $ with each $ {\mathsf P} ( A _ {j} ) $ either equal to some $ p $, $ 0 \leq p \leq 1 $, or zero. On the other hand, it was proved in [a6] that if (a1) is valid for the independent structures just mentioned, then (a1) is valid on an arbitrary probability space whatever the choice of the $ A _ {j} $. The above results reduce both (a1) and (a2) to polynomial inequalities. Amongst the several advantages of transforming (a1) and (a2) to polynomials, known as the method of polynomials for proving Bonferroni-type inequalities, is the following reduction formula: Assume one has found coefficients $ c _ {k} ( n ) = c _ {k} ( n,0 ) $ and $ d _ {k} ( n ) = d _ {k} ( n,0 ) $ such that (a2) holds for $ r = 0 $ and some set $ T = T _ {0} $. Then both (a1) and (a2) hold for arbitrary $ r $, $ 0 \leq r \leq n $, with the set $ T = T _ {r} $ obtained from $ T _ {0} $ by adding $ r $ to each of its elements, and the coefficients of $ S _ {k + r,n + r} $ in (a1) and (a2) satisfy the relations

$$ c _ {k + r} ( n + r,r ) = \left ( \begin{array}{c} {k + r} \\ r \end{array} \right ) c _ {k} ( n ) , $$

$$ d _ {k + r} ( n + r,r ) = \left ( \begin{array}{c} {k + r} \\ r \end{array} \right ) d _ {k} ( n ) . $$

Upon replacing $ \left ( \begin{array}{c} {k + r} \\ r \end{array} \right ) $ by $ \left ( \begin{array}{c} {k + r - 1} \\ {r - 1} \end{array} \right ) $ in the right-hand sides above, one obtains $ a _ {k + r} ( n + r,r ) $ and $ b _ {k+r} ( n + r,r ) $, respectively.

For modification ii) of the Bonferroni inequalities there are very general and very effective graph sieves. Choose an arbitrary graph with vertices $ \{ 1 \dots n \} $( that is, the edge set is chosen in an arbitrary manner). Now define $ S _ {k} ^ {*} $ by deleting many terms from $ S _ {k} $; in fact, retain in $ S _ {k + r} ^ {*} $ only those terms of $ S _ {k + r} $ for which $ ( i _ {1} \dots i _ {k + r} ) $ has no edge from the underlying graph just chosen if $ k $ is even, and allow, for $ k $ odd, a well-defined number of edges in $ ( i _ {1} \dots i _ {k + r} ) $, the number of which may depend on $ r $ but not on $ k $. Define $ S _ {k} ^ {* *} $ in a similar manner, but interchange the role of "even" and "odd" . It is proved in [a8], for $ r = 0 $, and in [a1], for arbitrary $ r $, that the Bonferroni lower bounds remain valid with $ S _ {k} ^ {*} $, whilst the upper bounds remain valid with $ S _ {k} ^ {* *} $. These new bounds found many applications in random set selections, information theory, and extreme value theory; see [a5]. The basic idea of constructing $ S _ {k} ^ {*} $ and $ S _ {k} ^ {* *} $ is similar to Brun's method in number theory (cf. Brun sieve).

Extension of Bonferroni and Bonferroni-type inequalities have also been studied in multivariate settings. Here, multivariate means that one faces several sequences of events, and one establishes linear bounds on the joint distribution of the numbers counting the occurrences in each sequence of events. Bounds are again in terms of binomial moments. Such multivariate studies are quite new; see [a3]. For a full coverage of the known (1996) multivariate bounds, see [a5].

References

[a1] J. Galambos, "On the sieve methods in probability theory. I" Studia Sci. Math. Hung. , 1 (1966) pp. 39–50
[a2] J. Galambos, "Bonferroni inequalities" Ann. of Probab. , 5 (1977) pp. 577–581
[a3] "Probability theory and applications" J. Galambos (ed.) I. Kátai (ed.) , Kluwer Acad. Publ. (1992)
[a4] J. Galambos, R. Mucci, "Inequalities for linear combination of binomial moments" Publ. Math. Debrecen , 27 (1980) pp. 263–269
[a5] J. Galambos, I. Simonelli, "Bonferroni-type inequalities with applications" , Springer (1996)
[a6] J. Galambos, I. Simonelli, "An extension of the method of polynomials and a new reduction formula for Bonferroni-type inequalities" Statistics and Probability Lett. , 28 (1996) pp. 147–151
[a7] S.M. Kwerel, "Most stringent bounds on aggregated probabilities of partially specified dependent probability systems" J. Amer. Statist. Assoc. , 70 (1975) pp. 472–479
[a8] A. Rényi, "A general method for proving theorems of probability theory and some of its applications" , Selected papers of Alfréd Rényi , 2 , Akad. Kiadó (1976) (Original (in Hungarian): MTA III Oszt. Közl. 11 (1961), 79–105)
How to Cite This Entry:
Bonferroni-type inequalities. Encyclopedia of Mathematics. URL: http://encyclopediaofmath.org/index.php?title=Bonferroni-type_inequalities&oldid=44398
This article was adapted from an original article by J. Galambos (originator), which appeared in Encyclopedia of Mathematics - ISBN 1402006098. See original article