mathematics

A gentle introduction to gradient descent thru linear regression

In this article we will briefly analyze one simple statistical tool that allows us to model a slice of reality by trying to assess, based on a set of Observations, how we can leverage a set of Variables to properly generate a model that will predict or forecast a behavior, by inferring about the variables relations and mutual influence. This statistical tool is called Linear Regression. In order to help minimizing the errors associated with the prediction model, optimizing it to better represent reality, we will also be briefly showing a simple application of Gradient Descent optimization algorithm.