Prediction with Model-Based Neutrality

Kazuto FUKUCHI; Toshihiro KAMISHIMA; Jun SAKUMA

doi:10.1587/transinf.2014EDP7367

Prediction with Model-Based Neutrality

Kazuto FUKUCHI, Toshihiro KAMISHIMA, Jun SAKUMA

Full Text Views

0

Cite this

Summary :

With recent developments in machine learning technology, the predictions by systems incorporating machine learning can now have a significant impact on the lives and activities of individuals. In some cases, predictions made by machine learning can result unexpectedly in unfair treatments to individuals. For example, if the results are highly dependent on personal attributes, such as gender or ethnicity, hiring decisions might be discriminatory. This paper investigates the neutralization of a probabilistic model with respect to another probabilistic model, referred to as a viewpoint. We present a novel definition of neutrality for probabilistic models, η-neutrality, and introduce a systematic method that uses the maximum likelihood estimation to enforce the neutrality of a prediction model. Our method can be applied to various machine learning algorithms, as demonstrated by η-neutral logistic regression and η-neutral linear regression.

Publication: IEICE TRANSACTIONS on Information Vol.E98-D No.8 pp.1503-1516

Publication Date: 2015/08/01

Publicized: 2015/05/15

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2014EDP7367

Type of Manuscript: PAPER

Category: Artificial Intelligence, Data Mining

Authors

Kazuto FUKUCHI
  University of Tsukuba
Toshihiro KAMISHIMA
  National Institute of Advanced Industrial Science and Technology (AIST)
Jun SAKUMA
  University of Tsukuba

Keyword

neutrality, fairness, discrimination, logistic regression, linear regression, classification, regression, social responsibility

Cite this

Copy

Kazuto FUKUCHI, Toshihiro KAMISHIMA, Jun SAKUMA, "Prediction with Model-Based Neutrality" in IEICE TRANSACTIONS on Information, vol. E98-D, no. 8, pp. 1503-1516, August 2015, doi: 10.1587/transinf.2014EDP7367.
Abstract: With recent developments in machine learning technology, the predictions by systems incorporating machine learning can now have a significant impact on the lives and activities of individuals. In some cases, predictions made by machine learning can result unexpectedly in unfair treatments to individuals. For example, if the results are highly dependent on personal attributes, such as gender or ethnicity, hiring decisions might be discriminatory. This paper investigates the neutralization of a probabilistic model with respect to another probabilistic model, referred to as a viewpoint. We present a novel definition of neutrality for probabilistic models, η-neutrality, and introduce a systematic method that uses the maximum likelihood estimation to enforce the neutrality of a prediction model. Our method can be applied to various machine learning algorithms, as demonstrated by η-neutral logistic regression and η-neutral linear regression.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2014EDP7367/_p

Copy

@ARTICLE{e98-d_8_1503,
author={Kazuto FUKUCHI, Toshihiro KAMISHIMA, Jun SAKUMA, },
journal={IEICE TRANSACTIONS on Information},
title={Prediction with Model-Based Neutrality},
year={2015},
volume={E98-D},
number={8},
pages={1503-1516},
abstract={With recent developments in machine learning technology, the predictions by systems incorporating machine learning can now have a significant impact on the lives and activities of individuals. In some cases, predictions made by machine learning can result unexpectedly in unfair treatments to individuals. For example, if the results are highly dependent on personal attributes, such as gender or ethnicity, hiring decisions might be discriminatory. This paper investigates the neutralization of a probabilistic model with respect to another probabilistic model, referred to as a viewpoint. We present a novel definition of neutrality for probabilistic models, η-neutrality, and introduce a systematic method that uses the maximum likelihood estimation to enforce the neutrality of a prediction model. Our method can be applied to various machine learning algorithms, as demonstrated by η-neutral logistic regression and η-neutral linear regression.},
keywords={},
doi={10.1587/transinf.2014EDP7367},
ISSN={1745-1361},
month={August},}

Copy

TY - JOUR
TI - Prediction with Model-Based Neutrality
T2 - IEICE TRANSACTIONS on Information
SP - 1503
EP - 1516
AU - Kazuto FUKUCHI
AU - Toshihiro KAMISHIMA
AU - Jun SAKUMA
PY - 2015
DO - 10.1587/transinf.2014EDP7367
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E98-D
IS - 8
JA - IEICE TRANSACTIONS on Information
Y1 - August 2015
AB - With recent developments in machine learning technology, the predictions by systems incorporating machine learning can now have a significant impact on the lives and activities of individuals. In some cases, predictions made by machine learning can result unexpectedly in unfair treatments to individuals. For example, if the results are highly dependent on personal attributes, such as gender or ethnicity, hiring decisions might be discriminatory. This paper investigates the neutralization of a probabilistic model with respect to another probabilistic model, referred to as a viewpoint. We present a novel definition of neutrality for probabilistic models, η-neutrality, and introduce a systematic method that uses the maximum likelihood estimation to enforce the neutrality of a prediction model. Our method can be applied to various machine learning algorithms, as demonstrated by η-neutral logistic regression and η-neutral linear regression.
ER -