Machine Learning Deep Learning Intermediate 2 min read

What is underfitting?

When a machine learning model is too simple to capture the underlying patterns in the data

underfitting explained in plain English

Underfitting occurs when a model is not complex enough to learn from the training data, resulting in poor performance on both the training and test data

Analogy

Imagine trying to draw a simple straight line to fit a complex curve - the line will not be able to capture the nuances of the curve, just like an underfitting model cannot capture the patterns in the data

Example

A company trying to predict customer churn using only a single feature, such as age, may experience underfitting because the relationship between age and churn is complex and influenced by many other factors

How is underfitting used?

Underfitting is often identified by high error rates on both the training and test data, and can be addressed by increasing the complexity of the model or adding more features to the data

Common misconceptions about underfitting

Underfitting is often confused with overfitting, but while overfitting occurs when a model is too complex, underfitting occurs when a model is too simple

History

The concept of underfitting has been around since the early days of machine learning, but it has become more prominent with the increasing use of complex models and large datasets

Also known as

bias high bias