Home Wikituition Browse all terms Categories
Random term
Mathematical & Computational Physics

Gradient Descent

An iterative first-order optimization algorithm for finding local minima of a differentiable scalar objective function by taking steps proportional to the negative gradient.

Governing formula x_(k+1) = x_k - α ∇f(x_k)
SI unit Iterative Optimization Method

In depth

Step size is controlled by learning rate parameter α. Used across computational physics and machine learning to minimize loss energy landscapes.

Examples in the real world

Finding ground state energy configurations in molecular modeling and training artificial neural networks.