Fairness in artificial intelligence for long-term equity

The National Science Foundation, in partnership with Amazon, has awarded Ohio State a Fairness in Artificial Intelligence (AI) grant for research on identifying and mitigating bias in AI and Machine Learning systems to achieve long-term equitable outcomes. Parinaz Naghizadeh, an assistant professor in Integrated Systems Engineering and Electrical and Computer Engineering is leading Ohio State's involvement in the program.

This will be a 3-year collaborative effort with colleagues at UC Santa Cruz (CS, lead), Michigan (ECE), and Purdue (CS), awarded at a budget of $1M ($240K to OSU). “We are hoping to provide analytical tools, validation and new datasets resulting from human subject experiments, and translation to use-cases in credit scoring and school admissions” says Naghizadeh.

Their proposal, titled "Fairness in Machine Learning with Human in the Loop", aims to understand the long-term fairness implications of automating decision-making using machine learning algorithms.

"We are specifically interested in accounting for human subjects whose behavior, participation incentives, qualification states, etc., will evolve over time when facing these algorithms, creating a feedback loop that further informs and complicates the design of fair AI” Naghizadeh says. 

The project “aims to understand the long-term impact of fair decisions made by automated machine learning algorithms via establishing an analytical, algorithmic, and experimental framework that captures the sequential learning and decision process, the actions and dynamics of the underlying user population, and its welfare” according to the proposal.