Write a python code and can run on Colab:

Build $5$ different regression models from the building blocks in this module, and measure their accuracy on a training, validation, and test set.

Use the same training and test split $($with same seed$),$ and use a validation set to do any hyperparameter tuning.

The code for splitting of test and training dataset :

from sklearn.model$\_$selection import train$\_$test$\_$split

X$\_$train, X$\_$test, y$\_$train, y$\_$test $=$ train$\_$test$\_$split$($X$,$ y$,$ test$\_$size$=0\mathrm{.}2,$ random$\_$state$=1)$

The code for dataset:

from sklearn import datasets, linear$\_$model

diabetes $=$ datasets.load$\_$diabetes$()$

# Use cross$-$validation for hyperparameter selection

Example:

regr $=$ linear$\_$model.ElasticNetCV$($cv$=5,$ random$\_$state$=1,$ l$1\_$ratio$=[\mathrm{.}1,\mathrm{.}5,\mathrm{.}7,\mathrm{.}9,\mathrm{.}95,\mathrm{.}99,1])$

Then, visualize the accuracy of your $5$ models as a barplot, with different$$subfigures$$for the accuracy on your $($a$)$ training, $($b$)$ validation, and $($c$)$ test data.

Last, discuss the results and whether the models are under$-$ or over$-$fitting.