next up previous print clean
Next: Masking missing data Up: Table of Contents Previous: Problem

Bilinear Solver

The Bilinear Operator is composite operator that is often used in missing data estimations at SEP. Furthermore, the Hessian of a vector valued function is a Bilinear Operator (see jam paper). All second order tensors are bilinear operators. A Bilinear Operator ${\bf A(x_1, x_2)}$ is an operator that is linear with respect to ${\bf x_1}$ when ${\bf x_2}$ is constant, and vice versa. A convolution is a bilinear operator with respect to its filter coefficients and its input data. Bilinear operators are better behaved then fully non-linear operators, e.g., their gradient is easily computed.


next up previous print clean
Next: Masking missing data Up: Table of Contents Previous: Problem
Stanford Exploration Project
3/8/1999