Stochastic convergence: Difference between revisions

From Citizendium
Jump to navigation Jump to search
imported>Ragnar Schroder
(Adding a paragraph to the header)
imported>Pat Palmer
mNo edit summary
(45 intermediate revisions by 5 users not shown)
Line 1: Line 1:
'''Stochastic convergence''' is a mathematical concept intended to formalize the idea that a sequence of essentially random or unpredictable events sometimes tends to settle into a pattern.
{{subpages}}


{{ dambigbox| Stochastic convergence | Convergence }}
'''Stochastic convergence''' is a mathematical concept intended to formalize the idea that a sequence of essentially random or unpredictable events sometimes is expected to settle into a pattern.


The pattern may for instance be
The pattern may for instance be
*[[convergence]] in the classical sense to a fixed value,  perhaps obtained from another random experiment.
*[[Convergence]] in the classical sense to a fixed value,  perhaps itself coming from a random event
*an increasing similarity of outcomes to what a purely deterministic function would produce
*An increasing similarity of outcomes to what a purely deterministic function would produce
*an increasing preference towards a certain outcome
*An increasing preference towards a certain outcome
*an increasing aversion for straying "far" away from a certain outcome
*An increasing "aversion" against straying far away from a certain outcome
*an increasing adherence to one particular probability distribution
 
*the series formed by calculating the [[expected value]] of the outcome's distance from a particular value may converge to 0
Some less obvious, more theoretical patterns could be
*the variability of the results grows smaller and smaller, i.e. the [[variance]] converges to 0
*That the probability distribution describing the next outcome may grow increasingly similar to a certain distribution
*That the series formed by calculating the [[expected value]] of the outcome's distance from a particular value may converge to 0
*That the variance of the [[random variable]] describing the next event grows smaller and smaller.




==Various possible modes of stochastic convergence==


Four different varieties of stochastic convergence are noted:
Four different varieties of stochastic convergence are noted:
Line 18: Line 24:
*Convergence in probability
*Convergence in probability
*Convergence in distribution
*Convergence in distribution
*Convergence in nth order mean
*Convergence in rth order mean
 
 
===Almost sure convergence===


This is the type of stochastic convergence that is most similar to [[pointwise convergence]] known from elementary [[real analysis]].


==Almost sure convergence==
{|align="right" cellpadding="10" style="background-color:lightblue; width:40%; border: 1px solid #aaa; margin:2px; font-size: 90%;"
|'''Examples of almost sure convergence.'''


===Example 1===
<b>Basic example 1</b>
Consider an animal of some short-lived species. We note the exact amount of food that this animal consumes day by day. This sequence of numbers will be unpredictable in advance, but we may be ''quite certain'' that one day the number will become zero, and will stay zero forever after.
Consider an animal of some short-lived species. We note the exact amount of food that this animal consumes day by day. This sequence of numbers will be unpredictable in advance, but we may be ''quite certain'' that one day the number will become zero, and will stay zero forever after.


===Example 2===
<b>Basic example 2</b>
Consider a man who starts tomorrow to toss seven coins once every morning.  Each afternoon, he donates a random amount of money to a certain charity.  The first time the result is all tails, however, he will stop permanently.
Consider a man who starts tomorrow to toss seven coins once every morning.  Each afternoon, he donates a random amount of money to a certain charity.  The first time the result is all tails, however, he will stop permanently.


Line 35: Line 47:
However,  when we consider ''any finite number'' of days,  there is a nonzero probability the terminating condition will not occur.
However,  when we consider ''any finite number'' of days,  there is a nonzero probability the terminating condition will not occur.


 
<b>Intermediate example</b>
===Example 3===
A business owner has two sources of income:  His business,  and interest from a large bank deposit with fixed interest and no withdrawal or deposits.
A business owner has two sources of income:  His business,  and interest from a large bank deposit with fixed interest and no withdrawal or deposits.


The business income varies unpredictably from month to month, while income from interest is predictable and given by a simple function f.
The business income varies unpredictably from month to month, while income from interest is predictable and given by a simple function f.


The income for month i can thus be modeled by a stochastic variable <math>U_i=X_i+f(i)</math>,  where <math>X_i</math> is the income from the  business.
The income for month i can thus be modeled by a random variable <math>U_i=X_i+f(i)</math>,  where <math>X_i</math> is the income from the  business.


Now assume <math>X_i</math> converges almost surely to 0 (history bears out that all businesses sooner or later fold up).  
Now assume <math>X_i</math> converges almost surely to 0 (history bears out that all businesses sooner or later fold up).  
Line 47: Line 58:
Then the total monthly income <math>U_i</math> has almost sure convergence to the function f(i).  
Then the total monthly income <math>U_i</math> has almost sure convergence to the function f(i).  


|}
===Informal description of almost sure convergence===
We are confronted with an infinite sequence of random experiments: Experiment 1, experiment 2, experiment 3 ... , where the outcome of each experiment will generate a real number. 
The random experiments will thus generate a  sequence of real numbers,  typically denoted ''x''<sub>1</sub>, ''x''<sub>2</sub>, ''x''<sub>3</sub>... .
If we have formulas available that describe the probabilities involved in each experiment,  then we may say something about the [[probability]] that this [[sequence]] will [[convergence|converge]] to a given value.


If this probability is 1, then the phenomenon of "almost sure convergence" is present.


===Formal definition===
Note that in [[Almost sure convergence|advanced treatments]] the outcomes are not restricted to real numbers.


Let <math>\scriptstyle X_0, X_1, ... </math> be an infinite sequence of [[stochastic variable|stochastic variables]] defined over a subset of R.  
====Formal definition====
 
Let '''X'''<sub>0</sub>, '''X'''<sub>1</sub>, '''X'''<sub>2</sub>... be an infinite sequence of [[random variable|random variables]] defined over a subset of R.  


Then the actual outcomes will be an ordinary sequence of real numbers.  
Then the actual outcomes will be an ordinary sequence of real numbers.  
Line 59: Line 80:
In more compact notation:
In more compact notation:


:If <math>P(\lim_{i \to \infty} X_i = a) = 1 </math> for some ''a'',  then the sequence has almost sure convergence to ''a''.
:If <math>P(\lim_{i \to \infty} X_i = a) = 1 </math> for some ''a'',  then the sequence has almost sure convergence to <math>a</math>.


Note that we may replace the real number a above by a real-valued function <math>f(i)</math> of i,  and obtain almost sure convergence to a function rather than a fixed number.
Note that we may replace the real number a above by a real-valued function <math>f(i)</math> of i,  and obtain almost sure convergence to a function rather than a fixed number.


The number a may also be the outcome of a random variable X.  In that case the compact but somewhat confusing notation <math>P(\lim_{i \to \infty} X_i = X) = 1 </math> is often used.


Commonly used notation: <math>X_i \stackrel{a.s.}{\rightarrow} a </math>, <math>X_i \stackrel{a.s.}{\rightarrow} X </math>.


==Convergence in probability==


===Example 1===


An absent-minded professor gets a job in an unfamiliar part of town.


The first time he walks to work,  he has difficulty finding his way,  and ends up several hours late.
===Convergence in probability===


The next few dozen times,  things generally improve,  although sometimes he manages to get hopelessly lost again.
The basic idea is that the probability of an "unusual" outcome becomes smaller and smaller as the sequence progresses.


As the months and years go by,  he gets to know the area very well,  and falls into a routine that make him more and more punctual,  although it may still happen occasionally that he is very late.


===Example 2===
{|align="right" cellpadding="10" style="background-color:lightblue; width:40%; border: 1px solid #aaa; margin:2px; font-size: 90%;"
We may keep tossing a die an infinite number of times and at every toss note the average outcome so far. The exact number thus obtained after each toss will be unpredictable,  but for a fair die,  it will tend to get closer and closer to the arithmetic average of 1,2,3,4,5 and 6,  i.e. 3.5.
|'''Illustration of convergence in probability.'''
{{Image|Convergenceinprobabilitysample1.gif|right|224px|Sequence of stochastic variables exhibiting typical behavior}}


Inspecting this graph,  you'll notice the behavior typical of convergence in probability. 


===Formal definition===
|}
Let <math>\scriptstyle X_0, X_1, ... </math> be an infinite sequence of [[stochastic variable|stochastic variables]] defined over a subset of R.  
 
 
 
 
 
====Formal definition====
Let <math>\scriptstyle X_0, X_1, ... </math> be an infinite sequence of [[random variable|random variables]] defined over a subset of R.  


If there exists a real number ''a'' such that <math>\lim_{i \to \infty} P( |X_i - a| > \varepsilon) = 0 </math>  
If there exists a real number ''a'' such that <math>\lim_{i \to \infty} P( |X_i - a| > \varepsilon) = 0 </math>  
Line 88: Line 115:




==Convergence in distribution==
Commonly used notation: <math>X_i \stackrel{P}{\rightarrow} a</math>.
===Example 1===
 
 
 
===Convergence in distribution===
With this mode of convergence,  we increasingly expect to see our next outcome in a sequence of random experiments becoming better and better modeled by a given [[probability distribution]].
 
{|align="right" cellpadding="10" style="background-color:lightblue; width:40%; border: 1px solid #aaa; margin:2px; font-size: 90%;"
|'''Examples of convergence in distribution.'''
 
<b>Basic example:</b>
The outcome from tossing a non-biased dice follows the [[uniform discrete distribution]].
The outcome from tossing a non-biased dice follows the [[uniform discrete distribution]].


Line 98: Line 134:
As the factory is improved,  the dices will be less and less loaded,  and the outcomes from tossing a newly produced dice will follow the desired distribution more and more closely.
As the factory is improved,  the dices will be less and less loaded,  and the outcomes from tossing a newly produced dice will follow the desired distribution more and more closely.


===Example 2===
<b>Intermediate example:</b>


Let <math>\scriptstyle X_n</math> be the result of flipping n unbiased coins,  and noting the fraction of heads.
Let <math>\scriptstyle X_n</math> be the result of flipping n unbiased coins,  and noting the fraction of heads.


<math>\scriptstyle X_1</math> will then follow the uniform discrete probability distribution with [[expected value of a stochastic variable|expected value]] <math>\mu=0.5</math> and [[variance of a stochastic variable|variance]] <math>\sigma^2=0.25</math>,  but as n grows larger,  <math>\scriptstyle X_n</math> will follow a distribution that gradually takes on more and more similarity to the [[normal distribution|gaussian distribution]] .
<math>\scriptstyle X_1</math> will then follow the uniform discrete probability distribution with [[expected value of a random variable|expected value]] <math>\mu=0.5</math> and [[variance of a random variable|variance]] <math>\sigma^2=0.25</math>,  but as n grows larger,  <math>\scriptstyle X_n</math> will follow a distribution that gradually takes on more and more similarity to the [[normal distribution|gaussian distribution]] .


Forming the stochastic sequence <math>\scriptstyle Z_n= \frac{ (X_n - \mu) }{\frac {\sigma} {\sqrt {n }}} </math>,  we find the variables <math>Z_n </math> becoming distributed more and more like the standard normal distribution as n increases.  
Forming the sequence <math>\scriptstyle Z_n= \frac{ (X_n - \mu) }{\frac {\sigma} {\sqrt {n }}} </math>,  we find the random variables <math>Z_n </math> becoming distributed more and more like the standard normal distribution as n increases.  


We then say the sequence <math>\scriptstyle Z_n</math> converges in distribution to the standard normal distribution.
We then say the sequence <math>\scriptstyle Z_n</math> converges in distribution to the standard normal distribution.
Line 110: Line 146:
(This convergence follows from the famous [[central limit theorem]]).
(This convergence follows from the famous [[central limit theorem]]).


===Formal definition===
|}
 
 
====Formal definition====
 
Given a random variable X with a [[cumulative distribution function]] F(x), let <math>X_i</math> be a sequence of random variables,  each with cumulative distribution function <math>F_i (x)</math>, respectively.
 
If <math>\scriptstyle \lim_{i \to \infty} F_i (x) = F(x)</math> for all x where F(x) is continuous,  then the sequence <math>X_i</math> of stochastic variables converges in distribution to the distribution of <math>X</math>.
 
Commonly used notation: <math>X_i \stackrel{L}{\rightarrow} X</math>.  One can also use the distribution directly,  so if f.i. X is normally distributed with mean 0 and variance 1,  one could write <math>X_i \stackrel{L}{\rightarrow} N(0,1)</math>.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
===Convergence in r-th order mean===
This is a rather "technical" mode of convergence. We essentially compute a sequence of real numbers,  one number for each random variable,  and check if this sequence is [[convergence|convergent]] in the ordinary sense.
 
{|align="right" cellpadding="10" style="background-color:lightblue; width:40%; border: 1px solid #aaa; margin:2px; font-size: 90%;"
|'''Examples of convergence in r-th order mean.'''
 
<b>Basic example:</b>
A newly built factory produces cans of beer.  The owners want each can to contain ''exactly'' a certain amount.


Knowing the details of the current production process,  engineers may compute the [[expected value|expected]] error in a newly produced can.


They are continuously improving the production process,  so as time goes by,  the expected error in a newly produced can tends to zero.


==Convergence in nth order mean==
This example illustrates convergence in first order mean.
===Example===
|}
===Formal definition===
 
If <math>\scriptstyle \lim_{n \to \infty} E(|X_n - a|^r ) =0</math>  for some real number a,  then {<math>X_n</math>} converges in rth order mean to a.
 
====Formal definition====
If <math>\scriptstyle \lim_{n \to \infty} E(|X_n - a|^r )=0</math>  for some real number a,  then {<math>X_n</math>} converges in r-th order mean to a.
 
Commonly used notation: <math>X_n \stackrel{L_r}{\rightarrow} a</math>.


==Relations between the different modes of convergence==
==Relations between the different modes of convergence==
*If a stochastic sequence has almost sure convergence,  then it also has convergence in probability.
*If a sequence of random variables has almost sure convergence,  then it also has convergence in probability.
*If a stochastic sequence has convergence in probability,  then it also has convergence in distribution.
*If a sequence of random variables has convergence in probability,  then it also has convergence in distribution.
*If a stochastic sequence has convergence in (n+1)th order mean, then it also has convergence in nth order mean (n>0).
*If a sequence of random variables has convergence in (r+1)-th order mean, then it also has convergence in r-th order mean (r>0).
*If a stochastic sequence has convergence in nth order mean, then it also has convergence in probability.
*If a sequence of random variables has convergence in rth order mean, then it also has convergence in probability.
 


== See also ==
== See also ==
*[[Stochastic differential equations]]
*[[Almost sure convergence]]
*[[Convergence in distribution]]
*[[Convergence in probability]]
*[[Convergence in r-th order mean]]
 


==Related topics==
==Related topics==
*[[Probability]]
*[[Probability]]
*[[Probability theory]]
*[[Probability theory]]
*[[Stochastic variable]]
*[[Random variable]]
*[[Differential equation|Differential equations]]
*[[Stochastic process]]
*[[Time series]]
*[[Stochastic differential equation]]
*[[Stochastic modeling]]
*[[Stochastic modeling]]


== References==
== References==
 
#P. Billingsley, ''Probability and Measure'' (2 ed.), ser. Wiley Series in Probability and Mathematical Statistics, Wiley, 1986.
#D. Williams, ''Probability with Martingales'', Cambridge : Cambridge University Press, 1991.
#E. Wong and B. Hajek, ''Stochastic Processes in Engineering Systems'', New York: Springer-Verlag,    1985.


== External links ==
== External links ==
 
#[http://www.probability.net Probability tutorial at Probability.net]
 
[[Category:CZ Live]]
[[Category:Mathematics Workgroup]]
[[Category:Physics Workgroup]]
[[Category:Chemistry Workgroup]]

Revision as of 18:57, 29 September 2020

This article is developing and not approved.
Main Article
Discussion
Related Articles  [?]
Bibliography  [?]
External Links  [?]
Citable Version  [?]
 
This editable Main Article is under development and subject to a disclaimer.
This article is about Stochastic convergence. For other uses of the term Convergence , please see Convergence (disambiguation).

Stochastic convergence is a mathematical concept intended to formalize the idea that a sequence of essentially random or unpredictable events sometimes is expected to settle into a pattern.

The pattern may for instance be

  • Convergence in the classical sense to a fixed value, perhaps itself coming from a random event
  • An increasing similarity of outcomes to what a purely deterministic function would produce
  • An increasing preference towards a certain outcome
  • An increasing "aversion" against straying far away from a certain outcome

Some less obvious, more theoretical patterns could be

  • That the probability distribution describing the next outcome may grow increasingly similar to a certain distribution
  • That the series formed by calculating the expected value of the outcome's distance from a particular value may converge to 0
  • That the variance of the random variable describing the next event grows smaller and smaller.


Various possible modes of stochastic convergence

Four different varieties of stochastic convergence are noted:

  • Almost sure convergence
  • Convergence in probability
  • Convergence in distribution
  • Convergence in rth order mean


Almost sure convergence

This is the type of stochastic convergence that is most similar to pointwise convergence known from elementary real analysis.

Examples of almost sure convergence.

Basic example 1 Consider an animal of some short-lived species. We note the exact amount of food that this animal consumes day by day. This sequence of numbers will be unpredictable in advance, but we may be quite certain that one day the number will become zero, and will stay zero forever after.

Basic example 2 Consider a man who starts tomorrow to toss seven coins once every morning. Each afternoon, he donates a random amount of money to a certain charity. The first time the result is all tails, however, he will stop permanently.

Let be the day by day amounts the charity receives from him.

We may be almost sure that one day this amount will be zero, and stay zero forever after that.

However, when we consider any finite number of days, there is a nonzero probability the terminating condition will not occur.

Intermediate example A business owner has two sources of income: His business, and interest from a large bank deposit with fixed interest and no withdrawal or deposits.

The business income varies unpredictably from month to month, while income from interest is predictable and given by a simple function f.

The income for month i can thus be modeled by a random variable , where is the income from the business.

Now assume converges almost surely to 0 (history bears out that all businesses sooner or later fold up).

Then the total monthly income has almost sure convergence to the function f(i).

Informal description of almost sure convergence

We are confronted with an infinite sequence of random experiments: Experiment 1, experiment 2, experiment 3 ... , where the outcome of each experiment will generate a real number. The random experiments will thus generate a sequence of real numbers, typically denoted x1, x2, x3... .

If we have formulas available that describe the probabilities involved in each experiment, then we may say something about the probability that this sequence will converge to a given value.

If this probability is 1, then the phenomenon of "almost sure convergence" is present.

Note that in advanced treatments the outcomes are not restricted to real numbers.

Formal definition

Let X0, X1, X2... be an infinite sequence of random variables defined over a subset of R.

Then the actual outcomes will be an ordinary sequence of real numbers.

If the probability that this sequence will converge to a given real number a equals 1, then we say the original sequence of stochastic variables has almost sure convergence to a.

In more compact notation:

If for some a, then the sequence has almost sure convergence to .

Note that we may replace the real number a above by a real-valued function of i, and obtain almost sure convergence to a function rather than a fixed number.

The number a may also be the outcome of a random variable X. In that case the compact but somewhat confusing notation is often used.

Commonly used notation: , .



Convergence in probability

The basic idea is that the probability of an "unusual" outcome becomes smaller and smaller as the sequence progresses.


Illustration of convergence in probability.
PD Image
Sequence of stochastic variables exhibiting typical behavior

Inspecting this graph, you'll notice the behavior typical of convergence in probability.



Formal definition

Let be an infinite sequence of random variables defined over a subset of R.

If there exists a real number a such that for all , then the sequence has convergence in probability to a.


Commonly used notation: .


Convergence in distribution

With this mode of convergence, we increasingly expect to see our next outcome in a sequence of random experiments becoming better and better modeled by a given probability distribution.

Examples of convergence in distribution.

Basic example: The outcome from tossing a non-biased dice follows the uniform discrete distribution.

Assume a new dice factory has just been built.

The first few dices come out quite biased, due to imperfections in the production process. The outcome from tossing any of them will follow a distribution markedly different from the desired uniform discrete distribution.

As the factory is improved, the dices will be less and less loaded, and the outcomes from tossing a newly produced dice will follow the desired distribution more and more closely.

Intermediate example:

Let be the result of flipping n unbiased coins, and noting the fraction of heads.

will then follow the uniform discrete probability distribution with expected value and variance , but as n grows larger, will follow a distribution that gradually takes on more and more similarity to the gaussian distribution .

Forming the sequence , we find the random variables becoming distributed more and more like the standard normal distribution as n increases.

We then say the sequence converges in distribution to the standard normal distribution.

(This convergence follows from the famous central limit theorem).


Formal definition

Given a random variable X with a cumulative distribution function F(x), let be a sequence of random variables, each with cumulative distribution function , respectively.

If for all x where F(x) is continuous, then the sequence of stochastic variables converges in distribution to the distribution of .

Commonly used notation: . One can also use the distribution directly, so if f.i. X is normally distributed with mean 0 and variance 1, one could write .














Convergence in r-th order mean

This is a rather "technical" mode of convergence. We essentially compute a sequence of real numbers, one number for each random variable, and check if this sequence is convergent in the ordinary sense.

Examples of convergence in r-th order mean.

Basic example: A newly built factory produces cans of beer. The owners want each can to contain exactly a certain amount.

Knowing the details of the current production process, engineers may compute the expected error in a newly produced can.

They are continuously improving the production process, so as time goes by, the expected error in a newly produced can tends to zero.

This example illustrates convergence in first order mean.


Formal definition

If for some real number a, then {} converges in r-th order mean to a.

Commonly used notation: .

Relations between the different modes of convergence

  • If a sequence of random variables has almost sure convergence, then it also has convergence in probability.
  • If a sequence of random variables has convergence in probability, then it also has convergence in distribution.
  • If a sequence of random variables has convergence in (r+1)-th order mean, then it also has convergence in r-th order mean (r>0).
  • If a sequence of random variables has convergence in rth order mean, then it also has convergence in probability.


See also


Related topics

References

  1. P. Billingsley, Probability and Measure (2 ed.), ser. Wiley Series in Probability and Mathematical Statistics, Wiley, 1986.
  2. D. Williams, Probability with Martingales, Cambridge : Cambridge University Press, 1991.
  3. E. Wong and B. Hajek, Stochastic Processes in Engineering Systems, New York: Springer-Verlag, 1985.

External links

  1. Probability tutorial at Probability.net