Probability theory MOC

Characteristic function

The characteristic function^[German charakteristische Funktion] πœ’(π‘˜) of a Real random variable 𝑋 βˆΌπ‘€ is the Fourier transform of the Probability density function 𝑀(π‘₯) or equivalently the Expectation of the function π‘’βˆ’π‘–π‘˜π‘‹1 prob

πœ’(π‘˜)=βŸ¨π‘’βˆ’π‘–π‘˜π‘‹βŸ©=F{𝑀}(π‘˜)=βˆ«βˆžβˆ’βˆžπ‘€(π‘₯)π‘’βˆ’π‘–π‘˜π‘₯𝑑π‘₯

which is a complex analogue to the moment-generating function. This describes the distribution of 𝑋 completely β€” the density function may be obtained using the reverse Fourier transform:

𝑀(π‘₯)=Fβˆ’1{πœ’}(π‘₯)=12πœ‹βˆ«βˆžβˆ’βˆžπœ’(π‘˜)π‘’π‘–π‘˜π‘₯π‘‘π‘˜

Using the Taylor series expansion of π‘’βˆ’π‘–π‘˜π‘‹ one obtains a further representation of πœ’(π‘˜) in terms of moments:

πœ’(π‘˜)=βˆžβˆ‘π‘›=0(βˆ’π‘–π‘˜)𝑛𝑛!βŸ¨π‘‹π‘›βŸ©


tidy | en | SemBr

Footnotes

  1. 2006, Statistische Mechanik, p. 5 ↩