demv 1.0.0

Debiaser for Multiple Variables, a model- and data- agnostic method to improve fairness in binary and multi-class classification tasks

DEMV : Debiaser for Multiple Variables

Table of contents

• Citation request
• Project structure
• General info
• DEMV class description
  • Attributes
  • Methods
  • Example usage

Citation request

Please cite our papers if you use DEMV in your experiments:

Giordano d’Aloisio, Andrea D’Angelo, Antinisca Di Marco, Giovanni Stilo, Debiaser for Multiple Variables to enhance fairness in classification tasks, Information Processing & Management, Volume 60, Issue 2, 2023, 103226, ISSN 0306-4573, https://doi.org/10.1016/j.ipm.2022.103226

@article{daloisio_debiaser_2023,
title = {Debiaser for Multiple Variables to enhance fairness in classification tasks},
journal = {Information Processing & Management},
volume = {60},
number = {2},
pages = {103226},
year = {2023},
issn = {0306-4573},
doi = {https://doi.org/10.1016/j.ipm.2022.103226},
url = {https://www.sciencedirect.com/science/article/pii/S0306457322003272},
author = {Giordano d’Aloisio and Andrea D’Angelo and Antinisca {Di Marco} and Giovanni Stilo},
keywords = {Machine learning, Bias and Fairness, Multi-class classification, Preprocessing algorithm, Equality},
}

d’Aloisio, G., Stilo, G., Di Marco, A., D’Angelo, A. (2022). Enhancing Fairness in Classification Tasks with Multiple Variables: A Data- and Model-Agnostic Approach. In: Boratto, L., Faralli, S., Marras, M., Stilo, G. (eds) Advances in Bias and Fairness in Information Retrieval. BIAS 2022. Communications in Computer and Information Science, vol 1610. Springer, Cham. https://doi.org/10.1007/978-3-031-09316-6_11

@inproceedings{d2022enhancing,
  title={Enhancing Fairness in Classification Tasks with Multiple Variables: A Data-and Model-Agnostic Approach},
  author={d’Aloisio, Giordano and Stilo, Giovanni and Di Marco, Antinisca and D’Angelo, Andrea},
  booktitle={International Workshop on Algorithmic Bias in Search and Recommendation},
  pages={117--129},
  year={2022},
  organization={Springer}
}

Project structure

This repository is organized as follows:

1. The source code of DEMV is in the demv folder.

General info

DEMV is a Debiaser for Multiple Variables that aims to increase Fairness in any given dataset, both binary and categorical, with one or more sensitive variables, while keeping the accuracy of the classifier as high as possible. The main idea behind the proposed method is that to enhance the classifier’s fairness during pre-processing effectively is necessary to consider all possible combinations of the values of the sensitive variables and the label’s values for the definition of the so-called sensitive groups.

We approach the problem by recursively identifying all the possible groups given by combining all the values of the sensible variables with the belonging label (class). Next, for each group, we compute its expected (𝑊𝑒𝑥𝑝) and observed (𝑊𝑜𝑏𝑠) sizes and look at the ratio among these two values. If 𝑊𝑒𝑥𝑝/𝑊𝑜𝑏𝑠 = 1, it implies that the group is fully balanced. Otherwise, if the ratio is less than one, the group size is larger than expected, so we must remove an element from the considered group accordingly to a chosen deletion strategy. Finally, if the ratio is greater than one, the group is smaller than expected, so we have to add another item accordingly to a generation strategy. For each group, we recursively repeat this balancing operation until 𝑊𝑒𝑥𝑝/𝑊𝑜𝑏𝑠 converge to one. It is worth noting that, in order to keep a high level of accuracy, the new items added to a group should be coherent in their values with the already existing ones.

The papers describing our work are available at:

• https://doi.org/10.1016/j.ipm.2022.103226
• http://dx.doi.org/10.1007/978-3-031-09316-6_11 (pdf).

DEMV class description

Attributes

• round_level : float

  Tolerance value to balance the sensitive groups

• debug : bool

  Prints w_exp/w_obs, useful for debugging

• stop : int

  Maximum number of balance iterations

• strategy: string

  Balancing strategy to use. Must be one of smote, adasyn and uniform (default is uniform)

• iter : int

  Maximum number of iterations

Methods

• __init__(self, round_level=None, debug=False, stop=-1, strategy='uniform')

    Creates a new DEMV instance
  
    Parameters
    ----------
    round_level : float, optional
        Tolerance value to balance the sensitive groups (default is None)
    debug : bool, optional
        Prints w_exp/w_obs, useful for debugging (default is False)
    stop : int, optional
        Maximum number of balance iterations (default is -1)
    strategy: string, optional
        Balancing strategy to use. Must be one of `smote`, `adasyn` and `uniform` (default is `uniform`)
  

• fit_transform(dataset: pandas.DataFrame, protected_attrs: list, label_name: str)

    Balances the dataset's sensitive groups
  
    Parameters
    ----------
    dataset : pandas.DataFrame
        Dataset to be balanced
    protected_attrs : list
        List of protected attribute names
    label_name : str
        Label name
  
    Returns
    -------
    pandas.DataFrame :
        Balanced dataset
  

• get_iters()

  Gets the maximum number of iterations
  
    Returns
    -------
    int:
        maximum number of iterations
  

Example usage

In the following we show an example usage of our algorithm:

from demv import DEMV
import pandas as pd

df = pd.read_csv('some_data.csv')
protected_attrs = ['s1','s2']
label = 'l'

demv = DEMV(round_level=1)
df_bal = demv.fit_transform(df, protected_attrs, label)
print('Maximum number of iterations: ',demv.get_iters())

Credits

The original paper was written by Giordano d'Aloisio, Giovanni Stilo, Antinisca di Marco and Andrea D'Angelo. This work is partially supported by Territori Aperti a project funded by Fondo Territori Lavoro e Conoscenza CGIL CISL UIL, by SoBigData-PlusPlus H2020-INFRAIA-2019-1 EU project, contract number 871042 and by “FAIR-EDU: Promote FAIRness in EDUcation institutions” a project founded by the University of L’Aquila. All the numerical simulations have been realized mostly on the Linux HPC cluster Caliban of the High-Performance Computing Laboratory of the Department of Information Engineering, Computer Science and Mathematics (DISIM) at the University of L’Aquila.

License

This work is licensed under AGPL 3.0 license.