Skewness and kurtosis

In modern data analysis, some high-order moments usually become useful, such as skewness and kurtosis.

  • Skewness:
    • Measures the asymmetry of the distribution.
    • skewness=E[(Xμσ)3]=:γ.
    • Gausian has skewness 0, that is, the distribution is symmetric.
  • Kurtosis:
    • Measure how heavy-tailed the distribution is.
    • kurtosis=E[(Xμσ)4]=k
    • Gaussian has kurtosis 3.
  • Empirical approximations
    • γ1NNn=1(Xnμσ)3

    • k1NNn=1(Xnμσ)4