Jef Jonkers

Background

About Jef Jonkers

As a PhD researcher in machine learning and statistics, I am driven to answer a fundamental question: How can we trust the predictions of AI systems? My research is dedicated to building this trust through rigorous uncertainty quantification, causal inference, and advanced probabilistic forecasting. My expertise is centered on conformal prediction, a robust, model-agnostic framework for creating statistically valid prediction intervals and distributions. I specialize in extending these methods to solve complex, real-world problems where standard assumptions often don't hold. This includes developing robust models for scenarios involving covariate shift, estimating causal effects, and forecasting high-dimensional systems.