# SGLB: Stochastic Gradient Langevin Boosting

arXiv: https://arxiv.org/abs/2001.07248 · HTML: https://mlsanity.com/paper/2001.07248/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Jan 2020.

## Built on by (most cited first, 12 of 12)

- [When to intervene? Prescriptive Process Monitoring Under Uncertainty and Resource Constraints](https://mlsanity.com/paper/2206.07745/) (2022)
- [Instance-Based Uncertainty Estimation for Gradient-Boosted Regression Trees](https://mlsanity.com/paper/2205.11412/) (2022)
- [Gradient Boosting Performs Gaussian Process Inference](https://mlsanity.com/paper/2206.05608/) (2022)
- [Which Tricks Are Important for Learning to Rank?](https://mlsanity.com/paper/2204.01500/) (2022)
- [Diffusion Boosted Trees](https://mlsanity.com/paper/2406.01813/) (2024)
- [SE(3)-Invariant Multiparameter Persistent Homology for Chiral-Sensitive Molecular Property Prediction](https://mlsanity.com/paper/2312.07633/) (2023)
- [High-Order Optimization of Gradient Boosted Decision Trees](https://mlsanity.com/paper/2211.11367/) (2022)
- [Variational Boosted Soft Trees](https://mlsanity.com/paper/2302.10706/) (2023)
- [Ito Diffusion Approximation of Universal Ito Chains for Sampling, Optimization and Boosting](https://mlsanity.com/paper/2310.06081/) (2023)
- [ClickTree: A Tree-based Method for Predicting Math Students' Performance Based on Clickstream Data](https://mlsanity.com/paper/2403.14664/) (2024)
- [Statistical Inference for Explainable Boosting Machines](https://mlsanity.com/paper/2601.18857/) (2026)
- [Conditioning Tree-Based Diffusions and Flows for Probabilistic Tabular Regression](https://mlsanity.com/paper/2607.28864/) (2026)

## Abstract (arXiv)

This paper introduces Stochastic Gradient Langevin Boosting (SGLB) - a powerful and efficient machine learning framework that may deal with a wide range of loss functions and has provable generalization guarantees. The method is based on a special form of the Langevin diffusion equation specifically designed for gradient boosting. This allows us to theoretically guarantee the global convergence even for multimodal loss functions, while standard gradient boosting algorithms can guarantee only local optimum. We also empirically show that SGLB outperforms classic gradient boosting when applied to classification tasks with 0-1 loss function, which is known to be multimodal.