Code for the experiments Section of the paper "From Soft Classifiers to Hard Decisions: How fair can we be?" (https://arxiv.org/pdf/1810.02003.pdf)
nishanthdikkala/postprocessing-deferrals
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
Low High +---------+ Didn't Reoffend |____|____| Reoffended | | | +---------+ This repository contains the code and plots used in the experiments section of the paper "From Soft Classifiers to Hard Decisions: How fair can we be?" (https://arxiv.org/pdf/1810.02003.pdf) Summary of the CSV files: 1. compas-scores-raw.csv: 28 X 60843 2. compas-scores-two-years-violent: 54 X 4743 3. compas-scores-two-years.csv: 53 X 7214 4. compas-scores.csv: 47 X 11757 Summary of the experiments performed: 1. Using deferrals on a single group to equalize the PDFs of the DOCS - deferral_equalize_pdf.R 2. Using deferrals on both groups via the min-PDF method to equalize the PDFs - deferral_min_pdf.R 3. Using deferrals with thresholds to equalize PPV and NPV across African-Americans and Caucasians simultaneously - deferral_thresholds.R
About
Code for the experiments Section of the paper "From Soft Classifiers to Hard Decisions: How fair can we be?" (https://arxiv.org/pdf/1810.02003.pdf)
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published