Data Scrubbing Operation: One-Hot Encoding

Quick overview

In data science, it’s common to have a mismatch or compatibility issue between the data and the algorithm. While the variable’s contents might be relevant, the algorithm might not read the data in its default form. Text-based categorical values, for example, can’t be parsed and mathematically modeled using general clustering and regression algorithms.

One quick remedy is to re-express categorical variables as a numeric categorizer. This can be performed using a common technique called one-hot encoding which converts categorical variables into binary form, represented as “1” or “0” or “True” or “False.”