Linear regression is the starter algorithm when it comes to machine learning. With the help of libraries like scikit learn, implementing multiple linear regression is hardly two or three lines of code. The dataset can also be handled easily with the help of pandas but I have tried to avoid that approach. In this article I have explained how to implement the algorithm using the classical approach of matrices using python. This approach focusses on implementing the algorithm straight from the pages of the book to code! The only package I have used is numpy, because our dataset deals with matrices, and numpy has many functions to efficiently handle them, without which we have to write a large number of loops, and that might still not be as effective as it would be with numpy. So let’s get started right away!
Linear regression with more than one input is called multiple linear regression or multivariate regression. In this implementation I have used the Real estate dataset which has several features and the algorithm tries to predict the price, which is the predictor.
#normal-equation #linear-regression-python #multiple-linearregression #machine-learning #numpy