In this repo, we made a regularized regression model that aimed to know further about our model is good or not good
Using boston.csv as a dataset, first, we have to split the data between train data and test data. Second, we have to know about VIF score using heatmap and drop the redundant features to avoid multicollinearity. Third step is find RMSE with 2 different model, which is ridge and lasso model, also find the coefficient. Last step is we have to train error and make interpretation about our model.
We can draw conclusion if the value of rmse in training error is not much different from the rmse in testing data, the model is acceptable.
pandas, numpy, scipy.stats, sklearn.model_selection, statsmodel, vif
Here are for the google.collab link