Difference between revisions of "Gradient descent"
From Wiki @ Karl Jones dot com
Karl Jones (Talk | contribs) |
Karl Jones (Talk | contribs) |
||
| Line 1: | Line 1: | ||
| − | '''Gradient descent''' is a [[First-order method|first-order]] [[Iterative method|iterative]] optimization [[algorithm]]. | + | '''Gradient descent''' is a [[First-order method|first-order]] [[Iterative method|iterative]] [[Mathematical optimization|optimization]] [[algorithm]]. |
== Description == | == Description == | ||
| Line 14: | Line 14: | ||
* [[First-order method]] | * [[First-order method]] | ||
* [[Iterative method]] | * [[Iterative method]] | ||
| + | * [[Mathematical optimization]] | ||
* [[Nelder–Mead method]] | * [[Nelder–Mead method]] | ||
* [[Preconditioning]] | * [[Preconditioning]] | ||
Revision as of 19:37, 12 October 2016
Gradient descent is a first-order iterative optimization algorithm.
Description
To find a local minimum of a function using gradient descent, one takes steps proportional to the negative of the gradient (or of the approximate gradient) of the function at the current point. If instead one takes steps proportional to the positive of the gradient, one approaches a local maximum of that function; the procedure is then known as gradient ascent.
Gradient descent is also known as steepest descent, or the method of steepest descent. Gradient descent should not be confused with the method of steepest descent for approximating integrals.
See also
- BFGS method
- Conjugate gradient method
- Delta rule
- First-order method
- Iterative method
- Mathematical optimization
- Nelder–Mead method
- Preconditioning
- Rprop
- Stochastic gradient descent
- Wolfe conditions
External links
- Gradient descent @ Wikipedia
