To prepare discrete categorical features for machine learning models, they are often transformed using one-hot encoding. For instance, consider a categorical feature like 'MSZoning' that assumes the values 'RL' and 'RM'. By dropping the original 'MSZoning' feature, two new indicator features (or dummy variables) are created: 'MSZoning_RL' and 'MSZoning_RM', which take values of either $$0$$ or $$1$$. If an example's original value for 'MSZoning' is 'RL', then its 'MSZoning_RL' feature is set to $$1$$ and its 'MSZoning_RM' feature is set to $$0$$.

Claude

A dummy variable, or indicator variable, is created to numerically represent categorical data. When a categorical feature has multiple possible discrete values, it is converted into a series of separate binary columns. For a given record, the column corresponding to its correct category is assigned a value of $$1$$, while all other dummy columns are set to $$0$$.

Dummy Variable

Dive into Deep Learning

A common imputation heuristic for missing categorical fields is to treat the absence of data itself as a distinct category. For example, if a column contains both valid string categories and NaN entries, it can be transformed into multiple binary dummy variables—where one specific column serves to indicate whether the original value was missing by using $$1$$ for missing and $$0$$ otherwise.

Categorical Imputation via Dummy Variables

The `pandas.get_dummies()` function is used to convert categorical variables into binary indicator columns. To apply categorical imputation, the `dummy_na=True` argument can be passed, which instructs the function to treat `NaN` as a valid category and generate a dedicated dummy column to explicitly flag missing entries.

Learn Before

Related