The National Science Foundation awarded Lu Zhang, Assistant Professor of Computer Science and Computer Engineering at the U of A, a prestigious Early Career Development Award to support his research on long-term equity in the sequential decision making.
The $597,185 prize over five years will be used to support Zhang’s research on Fair Machine Learning. Machine learning refers to the use and development of computer systems capable of learning and adapting without following explicit instructions by using algorithms and statistical models to analyze and make automatic decisions from patterns of data.
Zhang will use machine learning techniques to model how data is influenced by decisions in the wild. Then he will study how this influence could lead to just consequences by changing the decision-making strategy.
In his proposal, Zhang notes that “ensuring fairness for every decision does not guarantee long-term fairness, which poses a difficult and urgent problem for the fair machine learning community to achieve long-term fairness. term”.
Zhang uses the example of a bank making a loan to show how sequential decision making can impact long-term equity. If a bank grants a loan at an assigned interest rate, it can affect the recipient’s default risk and reduce their credit rating, which impacts their next loan application.
“If the bank’s decision causes a long-term decline in credit rating, it imposes a long-term negative effect on that person’s future decisions,” Zhang noted.
In this case, the objective is to ensure that the model initially used by the bank would not lead to unfair consequences, such as relying on historical data with biases. Other areas where sequential decision-making can lead to long-term unfairness may be job applications or college admissions.
Ultimately, Zhang hopes her work will help companies, organizations, and individuals better understand the benefits and risks of using machine learning in decision-making, whether their use imposes any consequences. unfair to certain groups of people and to improve the ability of domain users to comply with fairness. -related regulations.
To achieve this, Zhang will use Pearl’s Structural Causal Model, a mathematical framework that provides a general, formal computation for analyzing causal effects from observational data. He intends to propose universal formulations to measure long-term fairness, to develop learning algorithms to build fair decision models in offline and online learning contexts, and to study extensions to complex real-world situations.
“This project will bring transformative change to fair machine learning by significantly advancing the understanding of fundamental issues of fairness in dynamic contexts,” Zhang said, “illuminating the way forward for resolving conflicts between concepts of inconsistent fairness and contributing to the limited long-term fair machine learning knowledge base, which is imperative for many real-world applications.
The CAREER Awards are NSF’s most prestigious award for early-career faculty who have the potential to serve as academic role models in research and education and to lead advancements in their department or organization. The awards are for five years and include education and public awareness components. This award will help lay the foundation for Zhang’s career.
This is Zhang’s second NSF grant as a principal investigator. In October 2021, he received a grant of $484,828 from NSF’s Information and Intelligent Systems Division to support his research, “III: Small: Counterfactually Fair Machine Learning Through Causal Modeling.” The objective of this research was to reduce discrimination during intelligent machine learning in static systems. This CAREER award extends its previous research into the most important long-term consequences of automated decisions.
Warning: AAAS and EurekAlert! are not responsible for the accuracy of press releases posted on EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.