** Next:** LSJIMP Nonlinear Iterations
** Up:** The LSJIMP Inverse problem
** Previous:** Regularization 3: Crosstalk penalty

To effect the final step of LSJIMP, the estimation of the optimal set of
, we minimize a quadratic objective function which consists of
the sum of the weighted norms of the data residual [equation
()] and of the three model residuals [equations
(), (), and ()]:
| |
(12) |

and are scalars which balance the relative
weight of the three model residuals with the data residual. For the large scale
problems endemic to seismic imaging, the conjugate gradient method is a logical
choice to minimize .

** Next:** LSJIMP Nonlinear Iterations
** Up:** The LSJIMP Inverse problem
** Previous:** Regularization 3: Crosstalk penalty
Stanford Exploration Project

5/30/2004