Strong Wolfe Condition On Curvature
The Wolfe conditions, however, can result in a value for the step length that is not close to a minimizer of . If we modify the curvature condition to the following,
- iia)
then i) and iia) together form the so-called strong Wolfe conditions, and force to lie close to a critical point of .
The principal reason for imposing the Wolfe conditions in an optimization algorithm where is to ensure convergence of the gradient to zero. In particular, if the cosine of the angle between and the gradient,
is bounded away from zero and the i) and ii) hold, then .
An additional motivation, in the case of a quasi-Newton method is that if, where the matrix is updated by the BFGS or DFP formula, then if is positive definite ii) implies is also positive definite.
Read more about this topic: Wolfe Conditions
Famous quotes containing the words strong, wolfe and/or condition:
“Logicians may reason about abstractions. But the great mass of men must have images. The strong tendency of the multitude in all ages and nations to idolatry can be explained on no other principle.”
—Thomas Babington Macaulay (1800–1859)
“There had been a time on earth when poets had been young and dead and famous—and were men. But now the poet as the tragic child of grandeur and destiny had changed. The child of genius was a woman, now, and the man was gone.”
—Thomas Wolfe (1900–1938)
“No race has the last word on culture and on civilization. You do not know what the black man is capable of; you do not know what he is thinking and therefore you do not know what the oppressed and suppressed Negro, by virtue of his condition and circumstance, may give to the world as a surprise.”
—Marcus Garvey (1887–1940)