Optimization Algorithms

An optimization algorithm searches for parameters that improve an objective function. Gradient descent is a central method used to train many machine-learning models.

Japanese terms

1. Optimization algorithm — 最適化さいてきかアルゴリズム: A procedure for finding parameter values that improve or minimize an objective.
2. Objective function — 目的関数もくてきかんすう: A numerical function that an optimization process seeks to minimize or maximize.
3. Gradient descent — 勾配降下法こうばいこうかほう: An iterative optimization method that moves parameters in the direction that reduces an objective.

Related topics

Optimization is the numerical engine behind training in machine learning and deep learning.