Cynthia Rudin

Cynthia Rudin
Cynthia Rudin · CC BY 3.0 · source
Name	Cynthia Rudin
Nationality	American
Fields	Machine learning; Artificial intelligence; Interpretable models; Statistics; Criminal justice
Workplaces	Duke University; Massachusetts Institute of Technology; Columbia University
Alma mater	Duke University; Massachusetts Institute of Technology
Doctoral advisor	Ronald Rivest
Known for	Interpretable machine learning; Explainable artificial intelligence; Risk assessment critique
Awards	MacArthur Fellows Program; ACM SIGKDD; AAAI Fellows

Cynthia Rudin is an American computer scientist and statistician known for work on interpretable machine learning, explainable artificial intelligence, and algorithmic fairness. She has held faculty positions at several major research universities and has influenced debates about risk assessment tools in criminal justice, algorithmic transparency in healthcare, and interpretable models for scientific discovery. Her work bridges theoretical machine learning, practical applications, and public policy.

Early life and education

Rudin completed undergraduate studies at Duke University before pursuing graduate education at Massachusetts Institute of Technology, where she earned a Ph.D. under the supervision of Ronald Rivest. During her doctoral training she engaged with topics related to cryptography and statistical learning, connecting with research communities associated with ACM and IEEE. Her early mentors and collaborators included scholars from Columbia University and researchers active in the Association for Computing Machinery and the American Statistical Association.

Academic career and positions

Rudin has held academic appointments at Columbia University, where she developed courses and laboratories intersecting with research centers linked to New York University and Princeton University researchers. She later joined Duke University as a professor in departments that collaborate with units such as the Fuqua School of Business and the Durham research community. Her positions often involved interdisciplinary appointments spanning centers related to healthcare and public policy at institutions that partner with entities like Harvard University and MIT. She has delivered keynote talks at venues hosted by NeurIPS, ICML, AAAI, and KDD.

Research and contributions

Rudin’s research focuses on building transparent, provably interpretable models for high-stakes domains. She has proposed methods that contrast with opaque techniques produced by communities around deep learning, support vector machines, and ensemble methods often discussed at conferences such as NeurIPS and ICML. Her work includes algorithms for optimal decision trees, sparse scoring systems, and rule-based models used in domains related to medicine, criminal justice reform, and energy systems. She has published on limitations of proprietary risk assessment tools used in the United States criminal legal system and contributed to debates involving scholars from Harvard Law School, Stanford University, and Yale University about fairness and transparency. Rudin’s group has advanced open-source software and reproducible pipelines that interact with tools developed in communities tied to Python scientific ecosystems and projects affiliated with OpenAI and other industrial research groups.

Awards and honors

Rudin’s achievements have been recognized by major awards and fellowships. She received a fellowship from the MacArthur Fellows Program and has been named a Fellow of the Association for the Advancement of Artificial Intelligence and the Association for Computing Machinery in recognitions similar to honors given by IEEE. She has been awarded best paper and distinguished paper awards at conferences such as KDD and NeurIPS and received career awards from organizations akin to the National Science Foundation and national academies associated with United States science policy.

Selected publications and works

Rudin has authored influential articles and book chapters that appear in proceedings of NeurIPS, ICML, AAAI, and journals associated with Nature Machine Intelligence and Science Advances. Representative works include publications on optimal sparse decision trees, transparent scoring systems for healthcare triage, and critiques of algorithmic risk assessment used in pretrial settings that engage with literature from The New York Times and policy analyses from think tanks connected to Brookings Institution and RAND Corporation. Her group releases datasets and code that are used by researchers at institutions like Stanford University and Carnegie Mellon University.

Teaching and mentorship

Rudin has taught courses in interpretable machine learning and statistical learning theory that attract students from programs at Duke University, Columbia University, and partnering departments within Fuqua School of Business and engineering schools allied with MIT. She has supervised Ph.D. students who have taken faculty and industry roles at places such as Google Research, Microsoft Research, Amazon, and academic posts at Harvard University and Princeton University. Her mentoring emphasizes reproducible research practices and ethical deployment, aligning with curricular development at professional bodies like IEEE and ACM.

Public engagement and advocacy

Rudin is active in public discourse on algorithmic transparency and the social impacts of machine learning, engaging with policymakers at state and federal levels and participating in panels alongside experts from Harvard Kennedy School, Yale Law School, and advocacy groups connected to ACLU. She has testified and consulted regarding the evaluation of automated risk assessments used by prosecutors and courts in jurisdictions across the United States and collaborated with journalists from outlets including The New York Times, The Washington Post, and Nature. Rudin’s public-facing work aims to influence practice at technology companies and standards bodies such as ISO and professional societies like AAAI.

Category:Living people Category:Computer scientists Category:Women in computing