High cardinality refers to columns with values that are very uncommon or unique, example: email addresses, or user names. And low cardinality: refers to columns with few unique values, example: status flags, boolean values.
Card 1 of 259. Answer: High cardinality refers to columns with values that are very uncommon or unique, example: email addresses, or user names. And low cardinality: refers to columns with few unique values, example: status flags, boolean values.
About · Quizzes · History of AI · To suggest new questions or report an error send me a dm.
Cardinality is the number of distinct values a column holds. A high-cardinality column has many unique values, often close to one per row, like user IDs, email addresses or ZIP codes. A low-cardinality column has only a handful, like order status, gender or a yes/no flag.
It matters most when encoding categorical features. One-hot encoding a low-cardinality column is cheap and safe. One-hot encoding a high-cardinality one creates thousands of sparse columns, slows training and invites overfitting to rare values, so people reach for target encoding, count encoding, hashing or learned embeddings instead.
In databases the same word describes how selective a column is: indexing a high-cardinality column narrows a query a lot, while an index on a low-cardinality column rarely helps.
