Import the digit dataset from sklearn library

$1)$ Split the dataset with $75\%$ allocated for the training, use logistic regression, SVM$,$ Random forest, and decision Tree and get the score for for each Model

$2)$ Use the cross$\_$val$\_$score object with $5$ slots, and Measure the different scores for logistic regression, SVM$,$ Random forest, and decision Tree.

$3)$ Plot the confusion Matrix for the model with the best accuracy