• No products in the cart.

# 204.3.8 Practice : Validating the Tree

Link to the previous post : https://statinfer.com/204-3-7-building-a-decision-tree-in-python/

In the last post we built a decision tree and after plotting we explored the major characteristics of the tree.

In this post we will practice how to validate the tree.

### Tree Validation

• Find the accuracy of the classification for the tree model
#Tree Validation
predict1 = clf.predict(X)

from sklearn.metrics import confusion_matrix ###for using confusion matrix###
cm = confusion_matrix(y, predict1)
print (cm)

[[6370   38]
[ 648 4749]]

total = sum(sum(cm))
#####from confusion matrix calculate accuracy
accuracy = (cm[0,0]+cm[1,1])/total
accuracy

0.94188903007200342
• We can also use the .score() function to predict the accuracy in python from sklearn library.
• However, confusion matrix allows us to see the wrong classifications too that gives an intutive understanding.
clf.score(X,y)

0.94188903007200342

The next post is about the problem of overfitting the decision tree.
Link to the next post : https://statinfer.com/204-3-9-the-problem-of-overfitting-the-decision-tree/
21st June 2017
0 responses on "204.3.8 Practice : Validating the Tree"