Selecting informative conformal prediction sets with false coverage rate control

In our upcoming BDI Machine Learning seminar, we’re thrilled to welcome Dr Ruth Heller, Professor of Statistics from the Department of Statistics and Operations Research at Tel-Aviv University, Israel. We’re delighted to host Ruth in what promises to be a great talk!

Date: Friday 6 December
Time: 15:00 – 16:00
Talk title: Selecting informative conformal prediction sets with false coverage rate control
Location: BDI/OxPop Seminar room 1

Abstract:
In supervised learning, including regression and classification, conformal methods provide prediction sets for the outcome/label with finite sample coverage for any machine learning predictors. We consider here the case where such prediction sets come after a selection process. The selection process requires that the selected prediction sets be `informative’ in a well-defined sense. We consider both the classification and regression settings where the analyst may consider as informative only the sample with prediction sets small enough, excluding null values, or obeying other appropriate `monotone’ constraints. We develop a unified framework for building such informative conformal prediction sets while controlling the false coverage rate (FCR) on the selected sample. While conformal prediction sets after selection have been the focus of much recent literature in the field, the new introduced procedures, called InfoSP and InfoSCOP, are to our knowledge the first ones providing FCR control for informative prediction sets.
We show the usefulness of our resulting procedures on real and simulated data. This approach is applicable to neuroimaging, e.g., to provide prediction sets of medical (in particular neurological and psychiatric) diagnoses using the high dimensional individual functional MRI data.

Bio: Dr Heller is a Professor at the Department of Statistics and Operations Research at Tel-Aviv University, Israel, which she joined in 2011. She received her PhD in statistics from Tel-Aviv University in 2007, advised by Yoav Benjamini. Next, she spent two years at the Department of Statistic at the University of Pennsylvania, and two years at the Industrial Engineering and Management Faculty, Technion.

Her research interests are Multiple comparisons methods, Nonparamteric statistical tests. Observational studies, Post-selection inference, Replicability analysis and Conformal inference.