anjanragh/data-mining-on-propublica-compas-using-potentially-discriminatory--classification-association-rules-

In this project, I mined the post-processed subset of the ProPublica COMPAS dataset. I used SQL queries to perform data mining on it in PostGreSQL . The application took input support, confidence and protection thresholds. Its output were all PD-CARs that pass the specified thresholds P.S-You can read about the ProPublica COMPAS investigation here https:// www.propublica.org/article/machine-bias-risk-assessments-in-criminalsentencing and here https://www.propublica.org/article/how-we-analyzed-thecompas- recidivism-algorithm

Data-Mining-on-ProPublica-COMPAS-using-Potentially-Discriminatory--Classification-Association-Rules-

1. Load_PP.sql contains all the DMl script to create and store the data, to be mined.
2. PP.csv contains the Propublica Dataset.
3. The Compute_PP.sql file contains all the sql code required to mine the data from PP.csv file