Classifier comparison
The ROI framework can be run across different classifiers to see how higher precision and recall lead to higher ROI values. Note that the baseline classifier you created would have a total return and cost of 0 since both the true positives tp
and false positives fp
will be 0 by design. In this exercise, you will use the ROI framework to compare a logistic regression and decision tree classifier.
X_train
, y_train
, X_test
, y_test
are available in your workspace along with pandas
as pd
, numpy
as np
. LogisticRegression()
from sklearn.linear_model
is available as well.
This exercise is part of the course
Predicting CTR with Machine Learning in Python
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Create and fit classifier
clf = ____
y_pred = clf.____(X_train, y_train).____(X_test)
# Calculate total return, total spent, and ROI
r, cost = 0.2, 0.05
tn, fp, fn, tp = ____(y_test, y_pred).____
total_return = ____ * r
total_spent = (____ + ____) * cost
roi = total_return / total_spent
print("Total return: %s, Total spent: %s, ROI: %s" %(total_return, total_spent, roi))