Gradient Descent in Machine Learning

by Lleyton Fischer

Submitted : Fall 2020

Optimization algorithms are found in many disciplines and are used to find a better or best solution from a set of possible solutions. One area of interest that these algorithms are found in is the field of machine learning. Optimization algorithms are used to improve Machine Learning (ML) models and allow them to become more discerning of their input data sets. One such algorithm is the Gradient Descent algorithm, a simple and commonly used optimization algorithm that is found in many ML solutions. This paper investigates the Gradient Descent algorithm and its variants, their uses, and their basis in calculus. The Gradient Descent algorithm is an iterative algorithm that finds a minimum of a differentiable multi-variable function, which allows for finding more optimal parameters for a ML model.

[ Back ]