stochastic gradient descent
Stochastic gradient descent is a gradient descent method used for the minimization of an objective function.
All 1 facts, and where they come from
Facts
1 facts · 0 confirmed by two sources · 0 cited · 1 from one source
Where this comes from
Papalocal keeps its own record. It started from these sources; what each one said is kept here as it was copied, so a change on their side never changes this page by itself. Tap one to read its version.
What Wikidata says · copied November 19, 2026 · revision 2509291994
- Described as
- gradient descent method used for the minimization of an objective function
- Other names
- SGD, SGD optimizer
- What it is
- mathematical concept
- Wikipedias with a page
- 17
Wikidata item Q7617819 (wikidata.org/wiki/Q7617819), public domain (CC0). Copied 2026-11-19.
What Wikipedia says · copied November 19, 2026 · revision 1383580804
Stochastic gradient descent (often abbreviated SGD) is an iterative method for optimizing an objective function with suitable smoothness properties (e.g. differentiable or subdifferentiable). It can be regarded as a stochastic approximation of gradient descent optimization, since it replaces the actual gradient (calculated from the entire data set) by an estimate thereof, calculated from a randomly selected subset of the data. Especially in high-dimensional optimization problems this reduces the computational burden, achieving faster iterations in exchange for a lower convergence rate.
The basic idea behind stochastic approximation can be traced back to the Robbins–Monro algorithm of the 1950s. Today, stochastic gradient descent has become an important optimization method in machine learning.
From the Wikipedia article “Stochastic gradient descent” (en.wikipedia.org/wiki/Stochastic_gradient_descent), by its contributors, under CC BY-SA 4.0 (creativecommons.org/licenses/by-sa/4.0/). Copied 2026-11-19; not changed.
Confirmed Two sources agree. Cited One source, and it names where it got it. One source One source, without a reference of its own. Sources disagree The sources disagree. We show the best-supported value and keep theirs below.