Publication
TIC-TAC: A Framework for Improved Covariance Estimation in Deep Heteroscedastic Regression
Publications associées (39)
A Geometric Unification of Distributionally Robust Covariance Estimators: Shrinking the Spectrum by Inflating the Ambiguity Set
Daniel Kuhn, Yves Rychener, Viet Anh Nguyen
The state-of-the-art methods for estimating high-dimensional covariance matrices all shrink the eigenvalues of the sample covariance matrix towards a data-insensitive shrinkage target. The underlying shrinkage transformation is either chosen heuristically ...
2024Random matrix methods for high-dimensional machine learning models
In the rapidly evolving landscape of machine learning research, neural networks stand out with their ever-expanding number of parameters and reliance on increasingly large datasets. The financial cost and computational resources required for the training p ...
EPFL2024The two-point correlation function covariance with fewer mocks
We present FITCOV an approach for accurate estimation of the covariance of two-point correlation functions that requires fewer mocks than the standard mock-based covariance. This can be achieved by dividing a set of mocks into jackknife regions and fitting ...
Oxford2023Validation of semi-analytical, semi-empirical covariance matrices for two-point correlation function for early DESI data
We present an extended validation of semi-analytical, semi-empirical covariance matrices for the two-point correlation function (2PCF) on simulated catalogs representative of luminous red galaxies (LRGs) data collected during the initial 2 months of operat ...
OXFORD UNIV PRESS2023Positive Definite Completions and Continuous Graphical Models
This thesis concerns the theory of positive-definite completions and its mutually beneficial connections to the statistics of function-valued or continuously-indexed random processes, better known as functional data analysis. In particular, it dwells upon ...
EPFL2023Linear-Covariance Loss for End-to-End Learning of 6D Pose Estimation
Mathieu Salzmann, Yinlin Hu, Fulin Liu
Most modern image-based 6D object pose estimation methods learn to predict 2D-3D correspondences, from which the pose can be obtained using a PnP solver. Because of the non-differentiable nature of common PnP solvers, these methods are supervised via the i ...
Ieee Computer Soc2023Random Surface Covariance Estimation by Shifted Partial Tracing
The problem of covariance estimation for replicated surface-valued processes is examined from the functional data analysis perspective. Considerations of statistical and computational efficiency often compel the use of separability of the covariance, even ...
TAYLOR & FRANCIS INC2022CovNet: Covariance networks for functional data on multidimensional domains
Victor Panaretos, Soham Sarkar
Covariance estimation is ubiquitous in functional data analysis. Yet, the case of functional observations over multidimensional domains introduces computational and statistical challenges, rendering the standard methods effectively inapplicable. To address ...
WILEY2022Covariance Estimation for Random Surfaces beyond Separability
This thesis focuses on non-parametric covariance estimation for random surfaces, i.e.~functional data on a two-dimensional domain. Non-parametric covariance estimation lies at the heart of functional data analysis, andconsiderations of statistical and comp ...
EPFL2022Euclid : Effects of sample covariance on the number counts of galaxy clusters
Georges Meylan, Yi Wang, Richard Massey
Aims. We investigate the contribution of shot-noise and sample variance to uncertainties in the cosmological parameter constraints inferred from cluster number counts, in the context of the Euclid survey. Methods. By analysing 1000 Euclid-like light cones, ...
EDP SCIENCES S A2021