This activity aims to use the Decision Tree algorithm for $$pima$-$indians$-$diabetes$$ dataset.

Instructions:

Create a Decision Tree, train it to fit $75\%$ of the data, and test your model with the remaining $25\%.$

The following tasks can be performed:

$$ Import necessary tools

$$ Data exploration and preparation

$$ Creating and training the model

$$ The hyperparameter splitting criteria for categorical outputs are: $$gini$$ or $$entropy$.$ Entropy corresponds to the ID$3$ algorithm. The gini is the one used in the cart algorithm. Nevertheless, it is interesting to experiment with both splitting criteria.

$$ The hyperparameter $$max$\_$depth$$ can be varied to handle the overfitting. It fixes the depth of the tree.

$$ The hyperparameter $$min$\_$samples$\_$split$$ also can be varied; it corresponds to the minimum number of samples under which the node can be spitted.

$$ Tree visualization $($you can visualize the tree for different hyperparameters$)$

$$ Testing and measuring performance