Achieving accuracy and fairness in machine learning systems intended for use in social decision making is possible but designing those systems requires venturing off the simple and obvious paths.
Image credit: Falaah Arif Khan
CMU Researchers Dispel Theoretical Assumption About ML Trade-Offs in Policy Decisions
Carnegie Mellon University researchers are challenging a long-held assumption that there is a trade-off between accuracy and fairness when using machine learning to make public policy decisions.
As the use of machine learning has increased in areas such as criminal justice, hiring, health care delivery and social service interventions, concerns have grown over whether such applications introduce new or amplify existing inequities, especially among racial minorities and people with economic disadvantages. To guard against this bias, adjustments are made to the data, labels, model training, scoring systems and other aspects of the machine learning system. The underlying theoretical assumption is that these adjustments make the system less accurate.
A CMU team aims to dispel that assumption in a new study, recently published in Nature Machine Intelligence. Rayid Ghani, a professor in the School of Computer Science’s Machine Learning Department and the Heinz College of Information Systems and Public Policy; Kit Rodolfa, a research scientist in ML; and Hemank Lamba, a post-doctoral researcher in SCS, tested that assumption in real-world applications and found the trade-off was negligible in practice across a range of policy domains.
“You actually can get both. You don’t have to sacrifice accuracy to build systems that are fair and equitable,” Ghani said. “But it does require you to deliberately design systems to be fair and equitable. Off-the-shelf systems won’t work.”
Ghani and Rodolfa focused on situations where in-demand resources are limited, and machine learning systems are used to help allocate those resources. The researchers looked at systems in four areas: prioritizing limited mental health care outreach based on a person’s risk of returning to jail to reduce reincarceration; predicting serious safety violations to better deploy a city’s limited housing inspectors; modeling the risk of students not graduating from high school in time to identify those most in need of additional support; and helping teachers reach crowdfunding goals for classroom needs.
In each context, the researchers found that models optimized for accuracy — standard practice for machine learning — could effectively predict the outcomes of interest but exhibited considerable disparities in recommendations for interventions. However, when the researchers applied adjustments to the outputs of the models that targeted improving their fairness, they discovered that disparities based on race, age or income — depending on the situation — could be removed without a loss of accuracy.
Ghani and Rodolfa hope this research will start to change the minds of fellow researchers and policymakers as they consider the use of machine learning in decision making.
“We want the artificial intelligence, computer science and machine learning communities to stop accepting this assumption of a trade-off between accuracy and fairness and to start intentionally designing systems that maximize both,” Rodolfa said. “We hope policymakers will embrace machine learning as a tool in their decision making to help them achieve equitable outcomes.”
Original Article: Machine Learning Can Be Fair and Accurate
More from: Carnegie Mellon University
The Latest Updates from Bing News & Google News
Go deeper with Bing News on:
Machine learning fairness
- UK MHRA Publishes AI Regulatory Strategy
The AI White Paper set out five key principles for the regulation of AI: (i) safety, security and robustness; (ii) appropriate transparency and explainability; (iii) fairness ... further publications ...
- OU engineer receives NSF CAREER award for preeclampsia study
Talayah Razzaghi, an assistant professor of industrial and systems engineering at the University of Oklahoma, has been awarded a Faculty Early Career Development Program award from the National ...
- New mitigation framework reduces bias in classification outcomes
We use computers to help us make (hopefully) unbiased decisions. The problem is that machine-learning algorithms do not always make fair classifications if human bias is embedded in the data used to ...
- The bad and the ugly: AI is harmful, unreliable, and running out of data
Outperforming humans is one thing, but its rapid rise has meant that AI has created some problems for itself – and we're nervous.
- Clear guidelines needed for synthetic data to ensure transparency, accountability and fairness, study says
Clear guidelines should be established for the generation and processing of synthetic data to ensure transparency, accountability and fairness, a new study says.
Go deeper with Google Headlines on:
Machine learning fairness
[google_news title=”” keyword=”machine learning fairness” num_posts=”5″ blurb_length=”0″ show_thumb=”left”]
Go deeper with Bing News on:
Machine learning
- Machine learning harnessed toenhance high-power lasers
The experiment took place at the ELI Beamlines Facility in the Czech Republic, where the researchers utilised the state-of-the-art High-Repetition-Rate Advanced Petawatt Laser System (L3-HAPLS) to ...
- UChicago scientists use machine learning to turn cell snapshots dynamic
Pritzker School of Molecular Engineering study hopes to use machine learning to boost cancer, immunology research ...
- Snapchat Unveils New Augmented Reality And Machine Learning Tools For Brands And Advertisers
Discover Snapchat's latest AR & ML tools empowering brands & advertisers to engage audiences like never before!
- Machine Learning and Neural Network Can Be Effective Diagnostic Tools in MDS, Study Finds
The investigators said their new system makes it possible for researchers to leverage artificial intelligence even if they do not have expertise working with advanced software.
- Chrome address bar is becoming smarter with machine learning
Google is implementing machine learning in the latest Chrome update to make the address bar (omnibox) results more accurate.
Go deeper with Google Headlines on:
Machine learning
[google_news title=”” keyword=”machine learning” num_posts=”5″ blurb_length=”0″ show_thumb=”left”]
