Machine Learning - Linear Regression Model

Table of Contents

This article explains what Linear Regression Model is for Machine Learning.

What is Linear Regression Model? #

Linear Regression Model is a type of machine learning algorithm. It is predominantly used in predictive analysis and forecasting. The core idea behind a Linear Regression Model is to use a training set – a set of data used to guide the learning process – to train the model.

Notation #

\(x\) : feature, input variable

\(y\) : output variable, target variable

\(m\) : number of training examples

\((x, y)\) : single training example

\((x^{(i)}, y^{(i)})\) : \(i^{th}\) training example (\(1^{st}\), \(2^{nd}\), \(3^{rd}\), …)

flowchart TD A(training set) --> B(learning algorithm) B --> C(f) D(x) --> C C --> E(y hat)

x: feature f: model, function y hat : prediction

How to represent the model \(f\) ?

\(f_{w,b}(x)=wx+b\)

\(f(x)=wx+b\)

When the Linear Regression Model is used in the context of a single variable, it is often referred to as Univariate Linear Regression. This is a special case of the general Linear Regression Model where the output variable depends on a single input variable.