Skip to content

This project was carried out to determine the fairness of the compas algorithm in order to build a fairer model using responsible artificial intelligence techniques.

Notifications You must be signed in to change notification settings

SolomonAmaning/Compas-recidivism-analysis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 

Repository files navigation

Compas-recidivism-analysis

The COMPAS (Correctional Offender Management Profiling for Alternative Sanctions) recidivism algorithm is a widely used predictive tool that aims to assess the risk of recidivism among criminal offenders. It is designed to assist judges and parole boards in making informed decisions about pretrial release, sentencing, and parole. However, the algorithm has been a subject of controversy and scrutiny due to concerns about its fairness, transparency, and potential biases.

COMPAS analyzes a variety of factors, including the defendant's criminal history, age, gender, employment status, and substance abuse history, among others. Based on these inputs, it generates a risk score that predicts the likelihood of the individual committing another crime in the future. The scores are typically divided into categories such as low, medium, and high risk.

One of the main criticisms of the COMPAS algorithm is its potential for bias. Studies have shown that it tends to disproportionately label black defendants as higher risk and white defendants as lower risk, even when controlling for other relevant factors. This racial bias has raised concerns about the fairness of the decision-making processes that rely on COMPAS scores, as it may perpetuate existing racial disparities in the criminal justice system.

Another issue is the lack of transparency surrounding the inner workings of the COMPAS algorithm. The proprietary nature of the software and the limited disclosure of its specific details make it difficult for defendants and legal professionals to understand how the risk scores are calculated. This lack of transparency makes it challenging to identify and address any inherent biases or errors in the system.

The use of COMPAS in the criminal justice system has sparked numerous legal challenges and debates. Critics argue that relying heavily on an opaque algorithm to make crucial decisions about an individual's liberty raises concerns of due process and can perpetuate systemic inequalities. Proponents, on the other hand, contend that COMPAS provides valuable additional information to judges and parole boards, aiding in their decision-making process and potentially reducing recidivism rates.

This project was carried out to determine the fairness of the compas algorithm and in order build a fairer model using responsible artificial intelligence techniques.

My part of the project was to identify the proxy features. I used cosine similarity for this purpose.

About

This project was carried out to determine the fairness of the compas algorithm in order to build a fairer model using responsible artificial intelligence techniques.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published