ConceptConjugate Gradient / CG法
Conjugate Gradient Method
A classical iterative algorithm for solving symmetric positive definite linear systems. It constructs conjugate search directions incrementally to approach the solution, requiring only matrix-vector multiplications per iteration. Convergence speed depends on the condition number of the coefficient matrix.
Timeline (last 90 days)
Oct 14
In the deep learning context, exactly computing influence functions requires solving the product of the inverse Hessian matrix and the gradient vector, which is extremely costly, so in practice conjugate gradient methods or Hessian-vector products are commonly used for approximation.
Unverified50%