7. Evaluating Regressors

In [1]:

# Setup
from mlwpy import *
%matplotlib inline

diabetes = datasets.load_diabetes()

tts = skms.train_test_split(diabetes.data,
                            diabetes.target,
                            test_size=.25,
                            random_state=42)

(diabetes_train_ftrs, diabetes_test_ftrs,
 diabetes_train_tgt,  diabetes_test_tgt) = tts

We’ve discussed evaluation of learning systems and evaluation techniques specific to classifiers. Now, it is time to turn our focus to evaluating regressors. There are fewer ad-hoc techniques in evaluating regressors than in classifiers. For example, we don’t have confusion matrices and ROC curves, but we’ll see an interesting alternative in residual plots. Since we have some extra mental and physical space, we’ll spend a ...

Get Machine Learning with Python for Everyone now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.