Polished code, added a dataset, improved the readability

[T4ras123]

Changes made:

1. Added a iris.csv dataset to the directory so code in decision_tree_classification.ipynb can run out of the box.
2. Removed num_samples parameter from the get_best_split() function as it is not used.
3. Added f-strings and is not None in the if statements as a better practice.
4. Impoved redability by fixing spacing following pep8.