# Efficient Estimation of Regularized Tyler's M-Estimator Using Approximate LOOCV

arXiv: https://arxiv.org/abs/2505.24781 · HTML: https://mlsanity.com/paper/2505.24781/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 62 references that are in the map.
- First posted on arXiv 30 May 2025.

## Builds on (most cited first, 8 of 8)

- [Regularized Tyler's Scatter Estimator: Existence, Uniqueness, and Algorithms](https://mlsanity.com/paper/1407.3079/) (2014)
- [Regularized M-estimators of scatter matrix](https://mlsanity.com/paper/1405.2528/) (2014)
- [Shrinking the eigenvalues of M-estimators of covariance matrix](https://mlsanity.com/paper/2006.10005/) (2020)
- [Optimal shrinkage covariance matrix estimation under random sampling from elliptical distributions](https://mlsanity.com/paper/1808.10188/) (2018)
- [Multivariate outlier detection based on a robust Mahalanobis distance with shrinkage estimators](https://mlsanity.com/paper/1904.02596/) (2019)
- [Geodesic Convexity and Regularized Scatter Estimators](https://mlsanity.com/paper/1607.05455/) (2016)
- [Statistical analysis of latent generalized correlation matrix estimation in transelliptical distribution](https://mlsanity.com/paper/1305.6916/) (2013)
- [An a Priori Exponential Tail Bound for k-Folds Cross-Validation](https://mlsanity.com/paper/1706.05801/) (2017)

## Abstract (arXiv)

We consider the problem of estimating a regularization parameter, or a shrinkage coefficient α ∈ (0,1) for Regularized Tyler's M-estimator (RTME). In particular, we propose to estimate an optimal shrinkage coefficient by setting α as the solution to a suitably chosen objective function; namely the leave-one-out cross-validated (LOOCV) log-likelihood loss. Since LOOCV is computationally prohibitive even for moderate sample size n, we propose a computationally efficient approximation for the LOOCV log-likelihood loss that eliminates the need for invoking the RTME procedure n times for each sample left out during the LOOCV procedure. This approximation yields an O(n) reduction in the running time complexity for the LOOCV procedure, which results in a significant speedup for computing the LOOCV estimate. We demonstrate the efficiency and accuracy of the proposed approach on synthetic high-dimensional data sampled from heavy-tailed elliptical distributions, as well as on real high-dimensional datasets for object recognition, face recognition, and handwritten digit's recognition. Our experiments show that the proposed approach is efficient and consistently more accurate than other methods in the literature for shrinkage coefficient estimation.