Machine Learning with Spark - Second Edition

Feature hashing

Feature hashing is a technique to deal with high-dimensional data and is often used with text and categorical datasets where the features can take on many unique values (often many millions of values). In the previous chapters, we often used the 1-of-K encoding approach for categorical features, including text. While this approach is simple and effective, it can break down in the face of extremely high-dimensional data.

Building and using 1-of-K feature encoding requires us to keep a mapping of each possible feature value to an index in a vector. Furthermore, the process of creating the mapping itself requires at least one additional pass through the dataset and can be tricky to do in parallel scenarios. Up until now, we have ...