July 2017
Beginner to intermediate
486 pages
13h 49m
English
In this section, we will discuss the term underfitting. What is underfitting and how is it related to the bias-variance trade-off?
Suppose you train the data using any ML algorithm and you get a high training error. Refer to Figure 8.60:

The preceding situation, where we get a very high training error, is called underfitting. ML algorithms just can't perform well on the training data. Now, instead of a linear decision boundary, we will try a higher degree of polynomials. Refer to ...
Read now
Unlock full access