"maryam" wrote in message <kd487i$5pv$1@newscl01ah.mathworks.com>...
> "Torsten" wrote in message <kd3rgg$d1h$1@newscl01ah.mathworks.com>...
> > "maryam" wrote in message <kd3p99$3qu$1@newscl01ah.mathworks.com>...
> > > "Nasser M. Abbasi" wrote in message <kd34i6$p21$1@speranza.aioe.org>...
> > > > On 01/15/2013 01:47 AM, maryam wrote:
> > > > > Dear friens
> > > > > I would like calculate the expectated value (E(.)) of a 1-by-1000 matrix
> > > > > {E(x)=sigma x*f(x)}
> > > > > how do I that?
> > > >
> > > > Apply the definition of expected value?
> > > >
> > > > when you are not sure how to do something, always look up
> > > > the definition and go from there.
> > > >
> > > > http://en.wikipedia.org/wiki/Expected_value
> > > >
> > > > If you have problem with implementing the defintion, then show
> > > > what you tried and where you are having problem.
> > > >
> > > > good luck,
> > > >
> > > > --Nasser
> > > --------------------
> > > I know the the definition of expected value but I cant implement it.
> > > assume a 1-by-n arbitrary matrix, what's its matlab code?
> > > thanks
> >
> > A row of numbers does not have an expected value.
> > What you probably mean is an estimator for the expected value
> > of the random variable which produced the row of numbers.
> > Usually one takes
> > mean(A,2)
> > for this where A is your 1xn - matrix.
> >
> > Best wishes
> > Torsten.
> -----------------------------
> Dear Torsten
> thanks your reply
> In the part of my thesis I obtain a vector of random variables that is I=1xn - matrix.
> then I must to calculate "contrast (I)" , the definition of contrast is contrast(x)=E(x^2)/(E(x))^2
> and E(.) denoted the expectation operator.
> how could I calculate contrast (I)?
> thank you very much
The expected value of a set of random variables is based entirely on the distribution. For example the expected value of rolling a 6 sided dice is (6+1)/2 = 3.5
E(x^2) is not always safe to calculate directly
Now
Var(x) = E(x^2) - E(x)^2
so
contrast(x) = (Var(x) + E(x)^2)/E(x)^2
Find E(x) and Var(x) is trivial provided you know the distribution![]()
> "Torsten" wrote in message <kd3rgg$d1h$1@newscl01ah.mathworks.com>...
> > "maryam" wrote in message <kd3p99$3qu$1@newscl01ah.mathworks.com>...
> > > "Nasser M. Abbasi" wrote in message <kd34i6$p21$1@speranza.aioe.org>...
> > > > On 01/15/2013 01:47 AM, maryam wrote:
> > > > > Dear friens
> > > > > I would like calculate the expectated value (E(.)) of a 1-by-1000 matrix
> > > > > {E(x)=sigma x*f(x)}
> > > > > how do I that?
> > > >
> > > > Apply the definition of expected value?
> > > >
> > > > when you are not sure how to do something, always look up
> > > > the definition and go from there.
> > > >
> > > > http://en.wikipedia.org/wiki/Expected_value
> > > >
> > > > If you have problem with implementing the defintion, then show
> > > > what you tried and where you are having problem.
> > > >
> > > > good luck,
> > > >
> > > > --Nasser
> > > --------------------
> > > I know the the definition of expected value but I cant implement it.
> > > assume a 1-by-n arbitrary matrix, what's its matlab code?
> > > thanks
> >
> > A row of numbers does not have an expected value.
> > What you probably mean is an estimator for the expected value
> > of the random variable which produced the row of numbers.
> > Usually one takes
> > mean(A,2)
> > for this where A is your 1xn - matrix.
> >
> > Best wishes
> > Torsten.
> -----------------------------
> Dear Torsten
> thanks your reply
> In the part of my thesis I obtain a vector of random variables that is I=1xn - matrix.
> then I must to calculate "contrast (I)" , the definition of contrast is contrast(x)=E(x^2)/(E(x))^2
> and E(.) denoted the expectation operator.
> how could I calculate contrast (I)?
> thank you very much
The expected value of a set of random variables is based entirely on the distribution. For example the expected value of rolling a 6 sided dice is (6+1)/2 = 3.5
E(x^2) is not always safe to calculate directly
Now
Var(x) = E(x^2) - E(x)^2
so
contrast(x) = (Var(x) + E(x)^2)/E(x)^2
Find E(x) and Var(x) is trivial provided you know the distribution