Taylor expansion momentsΒΆ

In this page, we consider the Taylor expansion of a function. One way to evaluate the central dispersion (expectation and variance) of the variable \uY=h(\uX) is to use the Taylor expansion of the function h: \Rset^d \rightarrow \Rset^q at the mean point \vect{\mu} = \Expect{\uX}. Depending on the order of the Taylor expansion (classically first or second order), we get different approximations of the mean and variance of Y.

We use the notations introduced in Taylor Expansion.

In the remainder, let \Cov \uX be the covariance matrix of \uX, defined by:

\Cov \uX = \mat{C}

where \mat{C} \in \Rset^{n_X \times n_X} is the input covariance matrix:

c_{ij} = \Expect{\left(X_i - \Expect{X_i}\right)\left(X_j - \Expect{X_j} \right)}

with c_{ii} = \sigma_i^2.

Case 1: q=1, Y = h(\vect{X}) \in \RsetΒΆ

The second-order Taylor expansion of h at the point \ux = \vect{\mu} is:

y = h(\vect{\mu}) + \sum_{i = 1}^d \left( \frac{\partial h}{\partial x_i }\right)(\vect{\mu})(x_i-\mu_i) + \frac{1}{2} \sum_{i,j = 1}^d \left(\frac{\partial^2 h}{\partial x_i \partial x_j}\right)(\vect{\mu}) (x_i-\mu_i)(x_j-\mu_j) + o\left(\|\vect{x}\|^2\right).

The expectation and variance of the first-order expansion are:

\Expect{Y} \approx h(\vect{\mu})\\ \Var{Y} \approx \sum_{i=1}^{d} \sigma_i^2 \left(\left(\frac{\partial h}{\partial x_i} \right)(\vect{\mu}) \right)^2

The expectation of the second-order expansion is:

\Expect{Y} \approx h (\vect{\mu}) + \frac{1}{2} \sum_{i,j=1}^{d} c_{ij}\left(\frac{\partial^2 h} {\partial x_i \partial x_j}\right)(\vect{\mu}).

The second-order approximation of the variance is not implemented because it requires both the knowledge of higher order derivatives of h and the knowledge of moments of order strictly greater than 2 of the distribution of \uX.

Case 2: q>1, Y =(Y_1, \dots, Y_q) \in \Rset^qΒΆ

The second-order Taylor expansion of h = (h_1, \dots, h_q) at the point \ux = \vect{\mu} for each marginal function h_k is:

y_k = h_k(\vect{\mu}) + \sum_{i = 1}^d \left( \frac{\partial h_k}{\partial x_i }\right)(\vect{\mu}) (x_i-\mu_i)+ \frac{1}{2} \sum_{i,j = 1}^d \left( \frac{\partial^2 h_k}{\partial x_i \partial x_j}\right)(\vect{\mu})(x_i-\mu_i)(x_j-\mu_j) + o(\|\vect{x}\|^2).

where 1\leq k \leq q.

The expectation and covariance matrix of the first-order expansion are:

\Expect{\uY} \approx h(\vect{\mu})\\ \Cov \uY \approx \left( \sum_{i,j=1}^{d} c_{ij} \left( \frac{\partial h_k}{\partial x_i } \right)(\vect{\mu})\left( \frac{\partial h_\ell}{\partial x_j }\right)(\vect{\mu})\right)_{k, \ell}

The expectation of the second-order expansion is:

(\Expect{\uY})_k \approx \Expect{Y_k} \approx h_k(\vect{\mu}) + \frac{1}{2} \sum_{i,j=1}^{d} c_{ij}\left( \frac{\partial^2 h_k}{\partial x_i \partial x_j}\right)(\vect{\mu})

The second-order approximation of the variance is not implemented because it requires both the knowledge of higher order derivatives of h and the knowledge of moments of order strictly greater than 2 of the probability density function.