# Difference between revisions of "Eleisha's Segment 9: Characteristic Functions"

(9 intermediate revisions by the same user not shown) | |||

Line 3: | Line 3: | ||

1. Use characteristic functions to show that the sum of two independent Gaussian random variables is itself a Gaussian random variable. What is its mean and variance? | 1. Use characteristic functions to show that the sum of two independent Gaussian random variables is itself a Gaussian random variable. What is its mean and variance? | ||

+ | The characteristic function of the sum of independent random variables is the product of their characteristic functions. | ||

+ | |||

+ | <math> \text{The characteristic function of S is:} \phi_S(t) </math> | ||

+ | |||

+ | <math> X \sim \text{Normal}(\mu_1, \sigma_1) </math> | ||

+ | |||

+ | <math> Y \sim \text{Normal}(\mu_2, \sigma_2) </math> | ||

+ | |||

+ | So let S = X + Y (the sum of two independent Gaussian random variables) | ||

+ | |||

+ | <math> \phi_S(t) = \phi_X(t) \phi_Y(t) \text{ since X and Y are independent} </math> | ||

+ | |||

+ | <math> \phi_{X}(t) = e^{i \mu_1 t - \frac{1}{2} \sigma_1^2 t^2 } </math> | ||

+ | |||

+ | <math> \phi_{Y}(t) = e^{i \mu_2 t - \frac{1}{2} \sigma_2^2 t^2 } </math> | ||

+ | |||

+ | <math> \phi_S(t) = e^{i \mu_1 t - \frac{1}{2} \sigma_1^2 t^2 } \cdot e^{i \mu_2 t - \frac{1}{2} \sigma_2^2 t^2 } </math> | ||

+ | |||

+ | <math> \phi_S(t) = e^{i (\mu_1 + \mu_2) t - \frac{1}{2} (\sigma_1^2 + \sigma_2^2) t^2 } </math> | ||

+ | |||

+ | <math> \text{Notice that this characteristic function is just the characteristic function of a Normal with mean } (\mu_1 + \mu_2) \text{ and variance } (\sigma_1^2 + \sigma_2^2) </math> | ||

2. Calculate (don't just look up) the characteristic function of the Exponential distribution. | 2. Calculate (don't just look up) the characteristic function of the Exponential distribution. | ||

Line 8: | Line 29: | ||

<math> \phi_X(t) = \int_0^\infty e^{itx} \cdot \lambda e^{\lambda x} dx </math> | <math> \phi_X(t) = \int_0^\infty e^{itx} \cdot \lambda e^{\lambda x} dx </math> | ||

− | <math> \phi_X(t) = \lambda \int_0^\infty e^{ | + | <math> \phi_X(t) = \lambda \int_0^\infty e^{-(\lambda - it)x} dx </math> |

+ | |||

+ | <math> \phi_X(t) = \lambda\lim_{u \to \infty} \left[ \int_0^u e^{-(\lambda - it)x} dx \right] </math> | ||

+ | |||

+ | <math> \phi_X(t) = \lambda \lim_{u \to \infty} \left[ \frac{-1}{\lambda - it}\right] e^{-(\lambda - it)x}\Big|_0^u </math> | ||

+ | |||

+ | <math> \phi_X(t) = \lambda \left[ \lim_{u \to \infty} e^{-(\lambda - it)u} + \frac{1}{\lambda - it} \right] </math> | ||

+ | |||

+ | <math> \phi_X(t) = \lambda \left[ \frac{1}{\lambda - it}\right] </math> | ||

+ | |||

+ | <math> \phi_X(t) = \frac{\lambda}{\lambda - it} </math> | ||

<b>To Think About: </b> | <b>To Think About: </b> | ||

Line 16: | Line 47: | ||

2. Do you think that characteristic functions are ever useful computationally (that is, not just analytically to prove theorems)? | 2. Do you think that characteristic functions are ever useful computationally (that is, not just analytically to prove theorems)? | ||

− | <b>Back to</b> [[Eleisha Jackson]] | + | <b>Back to: </b> [[Eleisha Jackson]] |

## Latest revision as of 14:54, 22 February 2014

**To Calculate: **

1. Use characteristic functions to show that the sum of two independent Gaussian random variables is itself a Gaussian random variable. What is its mean and variance?

The characteristic function of the sum of independent random variables is the product of their characteristic functions.

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \text{The characteristic function of S is:} \phi_S(t) }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle X \sim \text{Normal}(\mu_1, \sigma_1) }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle Y \sim \text{Normal}(\mu_2, \sigma_2) }**

So let S = X + Y (the sum of two independent Gaussian random variables)

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \phi_S(t) = \phi_X(t) \phi_Y(t) \text{ since X and Y are independent} }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \phi_{X}(t) = e^{i \mu_1 t - \frac{1}{2} \sigma_1^2 t^2 } }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \phi_{Y}(t) = e^{i \mu_2 t - \frac{1}{2} \sigma_2^2 t^2 } }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \phi_S(t) = e^{i \mu_1 t - \frac{1}{2} \sigma_1^2 t^2 } \cdot e^{i \mu_2 t - \frac{1}{2} \sigma_2^2 t^2 } }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \phi_S(t) = e^{i (\mu_1 + \mu_2) t - \frac{1}{2} (\sigma_1^2 + \sigma_2^2) t^2 } }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \text{Notice that this characteristic function is just the characteristic function of a Normal with mean } (\mu_1 + \mu_2) \text{ and variance } (\sigma_1^2 + \sigma_2^2) }**

2. Calculate (don't just look up) the characteristic function of the Exponential distribution.

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \phi_X(t) = \int_0^\infty e^{itx} \cdot \lambda e^{\lambda x} dx }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \phi_X(t) = \lambda \int_0^\infty e^{-(\lambda - it)x} dx }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \phi_X(t) = \lambda\lim_{u \to \infty} \left[ \int_0^u e^{-(\lambda - it)x} dx \right] }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \phi_X(t) = \lambda \lim_{u \to \infty} \left[ \frac{-1}{\lambda - it}\right] e^{-(\lambda - it)x}\Big|_0^u }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \phi_X(t) = \lambda \left[ \lim_{u \to \infty} e^{-(\lambda - it)u} + \frac{1}{\lambda - it} \right] }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \phi_X(t) = \lambda \left[ \frac{1}{\lambda - it}\right] }**

**Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://wikimedia.org/api/rest_v1/":): {\displaystyle \phi_X(t) = \frac{\lambda}{\lambda - it} }**

**To Think About: **

1. Learn enough about contour integration to be able to make sense of Saul's explanation at the bottom of slide 7. Then draw a picture of the contours, label the pole(s), and show how you calculate their residues.

2. Do you think that characteristic functions are ever useful computationally (that is, not just analytically to prove theorems)?

**Back to: ** Eleisha Jackson