Three naive Bayes approaches for discrimination-free classification

Data Mining and Knowledge DiscoveryPublished 26 July 2010Open access

Toon Calders, Sicco Verwer

DOI10.1007/s10618-010-0190-x

Source URL

Citations774

SJR quartileQ1

SJR score1.02

SNIP1.88

View PDF

TL;DR

Three approaches for making the naive Bayes classifier discrimination-free are presented: modifying the probability of the decision being positive, training one model for every sensitive attribute value and balancing them, and adding a latent variable to the Bayesian model that represents the unbiased label and optimizing the model parameters for likelihood using expectation maximization.

Abstract

In this paper, we investigate how to modify the naive Bayes classifier in order to perform classification that is restricted to be independent with respect to a given sensitive attribute. Such independency restrictions occur naturally when the decision process leading to the labels in the data-set was biased; e.g., due to gender or racial discrimination. This setting is motivated by many cases in which there exist laws that disallow a decision that is partly based on discrimination. Naive application of machine learning techniques would result in huge fines for companies. We present three approaches for making the naive Bayes classifier discrimination-free: (i) modifying the probability of the decision being positive, (ii) training one model for every sensitive attribute value and balancing them, and (iii) adding a latent variable to the Bayesian model that represents the unbiased label and optimizing the model parameters for likelihood using expectation maximization. We present experiments for the three approaches on both artificial and real-life data.

Keywords

Computer Science

The foundations of cost-sensitive learning

1,835 Citations2001Charles Elkan

It is argued that changing the balance of negative and positive training examples has little effect on the classifiers produced by standard Bayesian and decision tree learning methods, and the recommended way of applying one of these methods is to learn a classifier from the training set and then to compute optimal decisions explicitly using the probability estimates given by the classifier.

Drug SafetyPrinciples of Data Mining

1,034 Citations2007David J. Hand

The book consists of three sections and provides a tutorial overview of the principles underlying data mining algorithms and their application, and shows how all of the preceding analysis fits together when applied to real-world data mining problems.

Discrimination-aware data mining

681 Citations2008Dino Pedreshi, Salvatore Ruggieri +1 more

This approach leads to a precise formulation of the redlining problem along with a formal result relating discriminatory rules with apparently safe ones by means of background knowledge, and an empirical assessment of the results on the German credit dataset.

Building Classifiers with Independency Constraints

452 Citations2009Toon Calders, Faisal Kamiran +1 more

This paper studies the classification with independency constraints problem: find an accurate model for which the predictions are independent from a given binary attribute and proposes two solutions and presents an empirical validation.

Toward scalable learning with non-uniform class and cost distributions: a case study in credit card fraud detection

448 Citations1998Philip K. Chan, Salvatore J. Stolfo

A multi-classifier meta-learning approach to address very large databases with skewed class distributions and non-uniform cost per error and empirical results indicate that the approach can significantly reduce loss due to illegitimate transactions.

Classifying without discriminating

402 Citations2009Faisal Kamiran, Toon Calders

This paper proposes a new classification scheme for learning unbiased models on biased training data based on massaging the dataset by making the least intrusive modifications which lead to an unbiased dataset and learns a non-discriminating classifier.

Measuring Discrimination in Socially-Sensitive Decision Records

175 Citations2009Dino Pedreschi, Salvatore Ruggieri +1 more

This work tackles the problem of determining, given a dataset of historical decision records, a precise measure of the degree of discrimination suffered by a given group in a given context with respect to the decision, and introduces a collection of quantitative measures of discrimination.

ACM Transactions on Knowledge Discovery from DataData mining for discrimination discovery

172 Citations2010Salvatore Ruggieri, Dino Pedreschi +1 more

This article formalizes the processes of direct and indirect discrimination discovery by modelling protected-by-law groups and contexts where discrimination occurs in a classification rule based syntax and proposes two inference models and provides automatic procedures for their implementation.

View PDF

Lecture notes in computer scienceNearest Neighbour Classification with Monotonicity Constraints

89 Citations2008Wouter Duivesteijn, Ad Feelders

A modified nearest neighbour algorithm for the construction of monotone classifiers from data is proposed that often outperforms standard kNN in problems where the monotonicity constraints are applicable.

Mining optimal decision trees from itemset lattices

66 Citations2007Siegfried Nijssen, Élisa Fromont

DL8, an exact algorithm for finding a decision tree that optimizes a ranking function under size, depth, accuracy and leaf constraints is presented and it is shown that optimal decision trees can be extracted from lattices of itemsets in linear time.

View PDF