Folk provides easy access to datasets that can be used to benchmark machine learning algorithms. The goal of Folk is to facilitate and encourage work on fair machine learning among R users.

The Folk package has three key features:

get_()The get_() functions provide easy access to data. Currently, there is only one get_() function, get_acs(), which provides access to the US Census Bureau’s American Community Survey’s Public Use Microdata Sample.
set_task()The set_task() function preprocesses data for pre-defined prediction tasks. Pre-defined tasks can be viewed with show_tasks().
new_task()The new_task() function allows users to create custom tasks. A custom task created via new_task() returns an object consistent with that returned by set_task().

Folk is inspired by the folktables Python package. For more information on folktables see Ding, Hardt, Miller, and Schmidt (2022), Retiring Adult: New Datasets for Fair Machine Learning. The pre-defined prediction tasks for the American Community Survey data are implementations of the tasks introduced in this paper.

Posted on:
November 14, 2023
1 minute read, 162 words
R package
See Also: