This repo is used for Multilineal Regression Practice
Profesor: Dr. Andrés Hernández Gutiérrez
Student: Alan Rocha
Test data need to be feature scaled with the same parameters (mean, standard variance) from the training data, in order to use this data for the prediction. Also a column of 1 needs to be added as the training dataset. With this is possible to obtain the prediction just multiplying the two matrices to obtain the result.
The table shows that the larger the learning rate, the less iterations and the faster the code runs, but the w values are less precise. This is because they perform "bigger jumps" with the largest learning rate.