Towards the end of "3 Introduction to general Machine Learning" a digit classification task was discussed.
Your task is to build a similar digit recongnizer, harnessing a DecisionTreeClassifier rather than a GaussianNB classifier. That includes:

1. Importing the dataset
2. Perform train-test-split
3. Instantiate and train the classifier
   • Find and tweak the hyperparameter to gain optimal results (try extreme values and see their influence)
4. Print the confusion matrix (from sklearn.metrics import confusion_matrix) to visualize the model's weaknesses.
5. Compare the result to the result achieved with the GaussianNB classifier.

Import the dataset

Reminder: you find it here: from sklearn.datasets import load_digits

Train-test-split; Classifier training

Confusion matrix