# Throwing Vines at the Wall: Structure Learning via Random Search

arXiv: https://arxiv.org/abs/2510.20035 · HTML: https://mlsanity.com/paper/2510.20035/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 78 references that are in the map.
- First posted on arXiv 22 Oct 2025.

## Builds on (most cited first, 23 of 23)

- [Copulas as High-Dimensional Generative Models: Vine Copula Autoencoders](https://mlsanity.com/paper/1906.05423/) (2019)
- [BOtied: Multi-objective Bayesian optimization with tied multivariate ranks](https://mlsanity.com/paper/2306.00344/) (2023)
- [Cross-Validation with Confidence](https://mlsanity.com/paper/1703.07904/) (2017)
- [Nonparametric estimation of simplified vine copula models: comparison of methods](https://mlsanity.com/paper/1701.00845/) (2017)
- [Semiparametric conformal prediction](https://mlsanity.com/paper/2411.02114/) (2024)
- [Generative Models for Simulating Mobility Trajectories](https://mlsanity.com/paper/1811.12801/) (2018)
- [Prediction based on conditional distributions of vine copulas](https://mlsanity.com/paper/1807.08429/) (2018)
- [Learning Vine Copula Models For Synthetic Data Generation](https://mlsanity.com/paper/1812.01226/) (2018)
- [Winners with Confidence: Discrete Argmin Inference with an Application to Model Selection](https://mlsanity.com/paper/2408.02060/) (2024)
- [From Dependence to Causation](https://mlsanity.com/paper/1607.03300/) (2016)
- [Model selection in sparse high-dimensional vine copula models with application to portfolio risk](https://mlsanity.com/paper/1801.09739/) (2018)
- [D-vine copula based quantile regression](https://mlsanity.com/paper/1510.04161/) (2015)
- [Nonparametric C- and D-vine based quantile regression](https://mlsanity.com/paper/2102.04873/) (2021)
- [Simplified vine copula models: state of science and affairs](https://mlsanity.com/paper/2410.16806/) (2024)
- [D-vine quantile regression with discrete variables](https://mlsanity.com/paper/1705.08310/) (2017)
- [Solving estimating equations with copulas](https://mlsanity.com/paper/1801.10576/) (2018)
- [Sequential Bayesian Model Selection of Regular Vine Copulas](https://mlsanity.com/paper/1512.00976/) (2015)
- [Selection of Sparse Vine Copulas in High Dimensions with the Lasso](https://mlsanity.com/paper/1705.05877/) (2017)
- [Representing sparse Gaussian DAGs as sparse R-vines allowing for non-Gaussian dependence](https://mlsanity.com/paper/1604.04202/) (2016)
- [Dependence Modeling in Ultra High Dimensions with Vine Copulas and the Graphical Lasso](https://mlsanity.com/paper/1709.05119/) (2017)
- [Black-Box Model Confidence Sets Using Cross-Validation with High-Dimensional Gaussian Comparison](https://mlsanity.com/paper/2211.04958/) (2022)
- [Growing simplified vine copula trees: improving Dissmann's algorithm](https://mlsanity.com/paper/1703.05203/) (2017)
- [Towards more realistic climate model outputs: A multivariate bias correction based on zero-inflated vine copulas](https://mlsanity.com/paper/2410.15931/) (2024)

## Built on by (most cited first, 1 of 1)

- [Bayesian model selection of vine copulas: a loss-based perspective](https://mlsanity.com/paper/2606.21512/) (2026)

## Abstract (arXiv)

Vine copulas offer flexible multivariate dependence modeling and have become widely used in machine learning. Yet, structure learning remains a key challenge. Early heuristics, such as Dissmann's greedy algorithm, are still considered the gold standard but are often suboptimal. We propose random search algorithms and a statistical framework based on model confidence sets, to improve structure selection, provide theoretical guarantees on selection probabilities and excess risk, as well as serve as a foundation for ensembling. Empirical results on real-world data sets show that our methods consistently outperform state-of-the-art approaches.