معرفی
Azadeh Khaleghi is a Professor of Statistics at ENSAE - CREST, where she conducts research at the intersection of statistical theory and machine learning. Her academic journey spans both engineering and mathematics, reflecting in her interdisciplinary research approach.
Professor Khaleghi earned her Ph.D. in Mathematics from Université de Lille I - INRIA Lille (France) and completed her undergraduate and master's studies in Electrical and Computer Engineering at the University of Toronto. This unique educational background bridges theoretical mathematics with practical engineering applications.
Her research program focuses on the mathematical foundations of statistics and machine learning, with particular emphasis on long-memory and mixing processes. She develops nonparametric methods for estimating mixing coefficients and modeling complex dependencies in time series data. Her work addresses the challenge of designing sequential decision-making algorithms that remain robust to long-range dependencies and non-stationarities. These theoretical contributions have implications across various domains where traditional statistical methods might fail due to complex dependency structures. Her techniques for understanding dependence in stochastic processes also inform broader areas of statistical learning, including algorithmic fairness.
At ENSAE, Professor Khaleghi teaches courses including Information Theory for Machine Learning, Statistical Modelling Seminar, and Statistics 1. These courses reflect her expertise in theoretical statistics and their applications to modern machine learning challenges.
Professor Khaleghi maintains active scholarly communication through her ORCID profile (0000-0001-8643-5416) and Google Scholar presence, facilitating collaboration and dissemination of her research findings. Her work at CREST contributes to the center's mission of advancing statistical methodology and its applications to economic and social sciences.
Her research group at CREST focuses on developing rigorous mathematical frameworks for understanding complex dependencies in data, with applications ranging from financial time series analysis to algorithmic fairness considerations in machine learning systems. The group maintains strong connections with both theoretical and applied research communities, fostering interdisciplinary collaboration.
