Description
1) Previously you created a model using the PISA dataset. Build a model again, this time…
a. (10 points) Use Ridge regression and present your model along with appropriate outputs.
i. Discuss how this technique handles multicollinearity.
ii. Evaluate the residual plots. Present the appropriate plots, describe them, and draw
appropriate conclusions. Note: to look at the residual plots you can – after selecting
variables with ridge regression – build a model using lm and plot the model.
b. (10 points) Use LASSO regression and present your model along with appropriate outputs.
i. LASSO is a form of feature selection. Discuss how it reduced the feature space.
c. (10 points) Are the two models the same? Explain.
2) REMISSION
a. (10 points) Download “remission” and create a logistic model to predict remission.
i. Present your model.
b. (5 points) Notice that you are using the glm function.
i. Explain how this differs from lm.
c. (5 points) Evaluate the model particularly the independent variables.
