Moment (mathematics)
From Free net encyclopedia
Current revision
- See also moment (physics).
The concept of moment in mathematics evolved from the concept of moment in physics. The nth moment of a real-valued function f(x) of a real variable about a value c is
- <math>\mu'_n=\int_{-\infty}^\infty (x - c)^n\,f(x)\,dx.</math>
The moments about zero are usually referred to simply as the moments of a function. Usually, except in the special context of the problem of moments below, the function will be a probability density function. The moments (about zero) of a probability density function f(x) are the expected values of Xn, the moments about its mean μ are called central moments; These describe the shape of the function, independently of translation.
If (lower-case) f is a probability density function, then the value integral above is called the nth moment of the probability distribution. More generally, if (capital) F is a cumulative probability distribution function of any probability distribution, which may not have a density function, then the nth moment of the probability distribution is given by the Riemann-Stieltjes integral
- <math>E(X^n)=\int_{-\infty}^\infty x^n\,dF(x),</math>
where X is a random variable that has this distribution.
When
- <math>E(|X^n|) = \int_{-\infty}^\infty |x^n|\,dF(x) = \infty,\,</math>
then the moment is said not to exist. If the nth moment about any point exists, so does (n − 1)th moment, and all lower-order moments, about every point.
Contents |
Significance of the moments
The first moment about zero, if it exists, is the expectation of X, i.e. the mean of the probability distribution of X, designated μ. In higher orders, the central moments are more interesting than the moments about zero.
The nth central moment of the probability distribution of a random variable X is
- <math>\mu_n=E((X-\mu)^n).\,</math>
The first central moment is thus 0; the second central moment is the variance, the square root of which is the standard deviation, σ. The normalised nth central moment is the nth central moment divided by σn; the nth moment of t = (x − μ)/σ. These normalised central moments are dimensionless quantities, which represent the distribution independently of any linear change of scale.
Skewness
The third central moment represents the lopsidedness of the distribution; any symmetric distribution will have a third central moment of zero. The normalised third central moment is called the skewness, often γ. A distribution that is skewed to the left (the tail of the distribution is on the left) will have a negative skewness. A distribution that is skewed to the right, the tail of the distribution is on the right, will have a positive skewness.
For distributions that are not too different from the normal (or "Gaussian") distribution, the median will be somewhere near μ − γσ/6; the mode about μ − γσ/2.
Kurtosis
The fourth central moment determines whether the distribution is tall and skinny or short and squat, compared to the normal distribution of the same variance. Since it is the expectation of a fourth power, the fourth central moment is always positive (except for a point distribution); the fourth central moment of a normal distribution is 3σ4.
The kurtosis κ is defined to be the normalized fourth central moment minus 3. (Equivalently, as in the next section, it is the fourth cumulant divided by the square of the variance.) Some authorities do not subtract three, but it is usually more convenient to have the normal distribution at the origin of coordinates. If a distribution has a peak at the mean and long tails, the fourth moment will be high and the kurtosis positive; and conversely.
The kurtosis can be positive without limit, but κ must be greater than or equal to γ2 − 2; equality only holds for binary distributions. For unbounded skew distributions not too far from normal, κ tends to be somewhere in the area of γ2 and 2γ2.
The inequality can be proven by considering
- <math>\operatorname{E} ((T^2 - aT)^2)</math>
where T = (X − μ)/σ. This is the expectation of a square, so it is non-negative whatever a is; on the other hand, it's a quadratic in a. Its discriminant must be non-positive, which gives the required relationship.
Cumulants
The first moment and the second and third unnormalized central moments are linear in the sense that if X and Y are independent random variables then
- <math>\mu_1(X+Y)=\mu_1(X)+\mu_1(Y)\,</math>
and
- <math>\operatorname{var}(X+Y)=\operatorname{var}(X)+\operatorname{var}(Y)</math>
and
- <math>\mu_3(X+Y)=\mu_3(X)+\mu_3(Y)\,</math>
(These can also hold for variables which aren't independent. The first always holds; if the second holds, the variables are called uncorrelated).
This is true because these moments are the first three cumulants; the fourth cumulant is the kurtosis times σ4.
All the cumulants are polynomials in the moments; so are the factorial moments. The central moments are polynomials in the moments about zero, and conversely.
Sample moments
Moments are often estimated based on the sample moments
- <math>\mu'_n \approx \frac{1}{N}\sum_{i = 1}^{N} X^n_i</math>
without estimating the probability distribution first.
Problem of moments
The problem of moments seeks characterizations of sequences { μ′n : n = 1, 2, 3, ... } that are sequences of moments of some function f.
See also
- Standardized moment
- Moment about the mean
- Cumulant
- Method of moments
- Normal distribution
- Binomial distribution
- Hamburger moment problem
- Hausdorff moment problem
- Stieltjes moment problem
- Taylor expansions for the moments of functions of random variables