The National Science Foundation has awarded Haewon Jeong, an assistant professor in the University of California Santa Barbara’s Electrical and Computer Engineering (ECE) Department, with an Early CAREER Award, which is the federal agency’s most highly regarded honor for junior faculty (Photo credit: UC Santa Barbara).
The award includes a five-year, $558,000 grant to support her research group.
Her project, titled “From Dirty Data to Fair Prediction: Data Preparation Framework for End-to-End Equitable Machine Learning,” aims to address bias in data and enhance ethical objectives within the data-preparation pipeline.
As part of her NSF project, Jeong seeks to investigate the causes of counterintuitive results and establish guidelines for data scientists on achieving an optimal demographic mixture.
She believes that the amount of noise in the data impacts how the data should be balanced, with noise representing inaccuracies such as people not answering surveys truthfully or encountering language barriers.
Jeong hypothesizes that the fairest and least biased mixture includes more data from the lowest noise level group.
She also highlights that data often contains missing values and different formats, necessitating standardization.
Jeong emphasizes the importance of supplying better examples and data to AI algorithms to result in more fair and ethical learning.
She also addresses the need to mitigate bias early in the process by researching how to substitute missing entries with new values without introducing more bias, a process called imputation.
Jeong’s ultimate goal is to develop a software library that data scientists and AI developers can use for fairness-aware data preparation.
The library would include fair-imputation methods, bias-flow measurement toolkit, and algorithms.
Furthermore, Jeong seeks to address gender disparities in AI by proposing an educational agenda to attract and retain talented female students in the field.
She aims to design and host the “Girls’ AI Bootcamp,” a program tailored to engage female high school students and introduce them to the opportunities within computer science and AI.