# Machine Learning for Multi-Output Regression: When should a holistic multivariate approach be preferred over separate univariate ones?

arXiv: https://arxiv.org/abs/2201.05340 · HTML: https://mlsanity.com/paper/2201.05340/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 3 of its 36 references that are in the map.
- First posted on arXiv 14 Jan 2022.

## Builds on (most cited first, 3 of 3)

- [ranger: A Fast Implementation of Random Forests for High Dimensional Data in C++ and R](https://mlsanity.com/paper/1508.04409/) (2015)
- [Travel Time Prediction using Tree-Based Ensembles](https://mlsanity.com/paper/2005.13818/) (2020)
- [Multivariate analysis of covariance when standard assumptions are violated](https://mlsanity.com/paper/1902.10195/) (2019)

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

- [SEGAA: A Unified Approach to Predicting Age, Gender, and Emotion in Speech](https://mlsanity.com/paper/2403.00887/) (2024)

## Abstract (arXiv)

Tree-based ensembles such as the Random Forest are modern classics among statistical learning methods. In particular, they are used for predicting univariate responses. In case of multiple outputs the question arises whether we separately fit univariate models or directly follow a multivariate approach. For the latter, several possibilities exist that are, e.g. based on modified splitting or stopping rules for multi-output regression. In this work we compare these methods in extensive simulations to help in answering the primary question when to use multivariate ensemble techniques.