Tree-Structured Synergy of Large Language Models and Bayesian Optimization for Efficient CASH

TL;DR

This paper introduces LB-MCTS, a novel tree-structured optimization framework combining Bayesian Optimization and Large Language Models to efficiently solve the combined algorithm selection and hyperparameter tuning problem in AutoML.

Contribution

The paper presents LB-MCTS, a new trajectory-structured method that synergizes BO and LLMs within a shared search state for improved AutoML performance.

Findings

LB-MCTS outperforms existing BO, LLM, and hybrid methods on 104 datasets.

The shared state enables effective collaboration between BO and LLMs.

Adaptive proposal shifting improves search efficiency and solution quality.

Abstract

To lower the expertise barrier in machine learning, the AutoML community has focused on the CASH problem, which jointly automates algorithm selection and hyperparameter tuning. While traditional methods like Bayesian Optimization (BO) struggle with cold-start issues, Large Language Models (LLMs) can mitigate these through semantic priors. However, existing LLM-based optimizers generalize poorly to high-dimensional, structured CASH spaces. In this paper, we propose LB-MCTS, a trajectory-structured optimization framework that uses a Monte Carlo Tree Search tree as a shared state for algorithm selection, hyperparameter refinement, and BO-LLM proposer synergy. Within this shared state, BO provides algorithm-specific surrogate modeling for quantitative search, while the LLM exploits path-aware selective memory to generate semantic proposals and reflections. As the surrogate model improves, a reliability-aware proposer policy adaptively shifts from LLM-driven to BO-driven proposals within a unified search trajectory. Experiments on 104 AMLB datasets demonstrate that LB-MCTS consistently outperforms BO-based, LLM-based, and hybrid baselines.