Tutorial leader
Professor Eyke Hüllermeier, Institute of Informatics, Munich Center for Machine Learning University of Munich (LMU)
This tutorial is part of the series on the thematic focus of uncertainty quantification (Objective 4: Machine learning with quantified uncertainty).
Abstract
This tutorial aims to provide an overview of uncertainty representation and quantification in machine learning, a topic that has received increasing attention in the recent past. The main focus is on novel approaches for distinguishing and representing so-called aleatoric and epistemic uncertainty. By the end of the tutorial, attendees will have a comprehensive understanding of the fundamental concepts and recent advances in this field.
Outline
The tutorial will focus on recent developments in the representation and quantification of uncertainty. It will start with a short introduction, followed by an explanation of concepts and approaches for first-oder and second-order representations, respectively.
Structure (tentative):
1. Introduction
2. First-order uncertainty representation
2.1 Representations based on probability distributions (calibrated probabilities, post-hoc procedures, training strategies, calibration tests, etc.)
2.2 Representations based on sets (conformal prediction, set-based utility maximization, etc.)
2.3 Extensions for regression (calibration, prediction intervals, conformal prediction, etc.)
3. Second-order uncertainty representation
3.1 Aleatoric and epistemic uncertainty (second-order representations via distributions and sets, etc.)
3.2 Bayesian inference (Gaussian processes, Bayesian neural networks, ensemble-based approximations, etc.)
3.3 Generalized loss minimization (evidential deep learning, second-order losses, etc. )
3.4 Uncertainty quantification (information-theoretic measures, dispersion measures, etc.)
Target audience
This tutorial aims to attract both researchers that are already active in one of the above domains, as well as researchers with little or no prior experience in uncertainty quantification. We will only assume a general familiarity with well-known machine learning techniques for classification and regression (neural networks, probabilistic models, risk minimization, etc.). Therefore, the tutorial is of interest to scholars from diverse subfields of machine learning and with different backgrounds.
This tutorial is open for attendance for everybody.
Practical
No prior registration is required.
Zoom link: https://uio.zoom.us/j/63272965032