= Gradient method
{wiki=Gradient_method}
The gradient method, often referred to as Gradient Descent, is an optimization algorithm used to minimize a function by iteratively moving towards the steepest descent as defined by the negative of the gradient. It is widely used in various fields, particularly in machine learning and deep learning for optimizing loss functions. \#\#\# Key Concepts 1. **Gradient**: The gradient of a function is a vector that points in the direction of the steepest increase of that function.
Back to article page