Linear-Regression-Implementation

This project contains the .data files for both the testing and training data. It also contains housing data which was used to test the implementation of the linear regression models on a real set of data. There are 2 python files contained within this project. The linearRegression.py file is a low level implementation of the linear regression model using the normal equations and gradient descent to fit the model to the training data. For both the normal equations and gradient descent methods, there are 5 polynomial regression models fit to the training data with the 5 models being 0-4th orders of polynomial regression. The mean squared error for each model is then computed on the test set to record how well each model fits the data. The housingDataLinearRegression.py file does essentially the same thing where a linear regression model is fit to the data using the normal equations. The program uses K-folds cross validation and finds the model which best fits the data based upon computing mean squared errors for each k-fold model.