How to Solve AI Bias

Robin Bradley

How to Solve AI Bias

Robin Bradley

School of Physics, Engineering & Computer Science

Research output: Contribution to conference › Paper

115 Downloads (Pure)

Abstract

Bias in AI is a topic that impacts machine learning and artificial intelligence technology that learns from datasets and its training data. While gender discrimination and chatbots showing bias have recently caught people’s attention and imagination, the overall area of how to correct and manage bias is in its infancy for business use. Further, little is known about how to solve bias in AI and how there could potent for malicious misuse at large scale. We explore this area and propose solutions to this problem.

Original language	English
Number of pages	7
Publication status	Published - 17 Feb 2020
Event	Women in Tech - Bias in AI: Publicis Media Women in Tech: Bias in Algorithms - Television Centre, London, United Kingdom Duration: 17 Feb 2020 → 17 Feb 2020

Workshop

Workshop	Women in Tech - Bias in AI
Country/Territory	United Kingdom
City	London
Period	17/02/20 → 17/02/20

Keywords

Artificial intelligence
AI artificial intelligence
bias
Machine Learning
Algorithm and data structure visualization
Algorithms
ethics and politics
Ethics Committees, Research
ethics; socially responsible investment
testing methodology

Access to Document

How to Solve AI Bias - by Robin Bradley v2
© 2020 The Author(s). This an open access work distributed under the terms of the Creative Commons Attribution Licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
Final published version, 246 KBLicence: CC BY

Cite this

@conference{f6b4a6f7a100421a83ee78fc0b6b018d,

title = "How to Solve AI Bias",

abstract = "Bias in AI is a topic that impacts machine learning and artificial intelligence technology that learns from datasets and its training data. While gender discrimination and chatbots showing bias have recently caught people{\textquoteright}s attention and imagination, the overall area of how to correct and manage bias is in its infancy for business use. Further, little is known about how to solve bias in AI and how there could potent for malicious misuse at large scale. We explore this area and propose solutions to this problem.",

keywords = "Artificial intelligence, AI artificial intelligence, bias, Machine Learning, Algorithm and data structure visualization, Algorithms, ethics and politics, Ethics Committees, Research, ethics; socially responsible investment, testing methodology",

author = "Robin Bradley",

note = "{\textcopyright} 2020 The Author(s). This an open access work distributed under the terms of the Creative Commons Attribution Licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.; Women in Tech - Bias in AI : Publicis Media Women in Tech: Bias in Algorithms ; Conference date: 17-02-2020 Through 17-02-2020",

year = "2020",

month = feb,

day = "17",

language = "English",

}

TY - CONF

T1 - How to Solve AI Bias

AU - Bradley, Robin

N1 - © 2020 The Author(s). This an open access work distributed under the terms of the Creative Commons Attribution Licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.

PY - 2020/2/17

Y1 - 2020/2/17

N2 - Bias in AI is a topic that impacts machine learning and artificial intelligence technology that learns from datasets and its training data. While gender discrimination and chatbots showing bias have recently caught people’s attention and imagination, the overall area of how to correct and manage bias is in its infancy for business use. Further, little is known about how to solve bias in AI and how there could potent for malicious misuse at large scale. We explore this area and propose solutions to this problem.

AB - Bias in AI is a topic that impacts machine learning and artificial intelligence technology that learns from datasets and its training data. While gender discrimination and chatbots showing bias have recently caught people’s attention and imagination, the overall area of how to correct and manage bias is in its infancy for business use. Further, little is known about how to solve bias in AI and how there could potent for malicious misuse at large scale. We explore this area and propose solutions to this problem.

KW - Artificial intelligence

KW - AI artificial intelligence

KW - bias

KW - Machine Learning

KW - Algorithm and data structure visualization

KW - Algorithms

KW - ethics and politics

KW - Ethics Committees, Research

KW - ethics; socially responsible investment

KW - testing methodology

M3 - Paper

T2 - Women in Tech - Bias in AI

Y2 - 17 February 2020 through 17 February 2020

ER -

How to Solve AI Bias

Abstract

Workshop

Keywords

Access to Document

Fingerprint

Cite this