LLM Agent for Hyper-Parameter Optimization

2025-06-18

AI Summary

Manual hyperparameter tuning for radio map–assisted UAV trajectory and communication algorithms suffers from low automation and suboptimal performance. Method: This paper proposes the first large language model (LLM)-based agent framework that employs an LLM as an autonomous hyperparameter optimizer. Integrating task profiling, Model Context Protocol (MCP), and particle swarm optimization (PSO) domain knowledge, the framework performs closed-loop search via iterative invocation of Warm-Start PSO with Crossover and Mutation (WS-PSO-CM) and supports autonomous termination. Contribution/Results: Experiments in UAV communication scenarios demonstrate that the LLM-optimized hyperparameters significantly improve the minimum achievable rate over manual heuristics and random search. The results validate the effectiveness and generalizability of the LLM agent in complex engineering optimization tasks.

Technology Category

• Search and Optimization: Learning to Search
• Multiagent Systems: Agent Communication
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Hyper-parameters are essential and critical for the performance of communication algorithms. However, current hyper-parameters tuning methods for warm-start particles swarm optimization with cross and mutation (WS-PSO-CM) algortihm for radio map-enabled unmanned aerial vehicle (UAV) trajectory and communication are primarily heuristic-based, exhibiting low levels of automation and unsatisfactory performance. In this paper, we design an large language model (LLM) agent for automatic hyper-parameters-tuning, where an iterative framework and model context protocol (MCP) are applied. In particular, the LLM agent is first setup via a profile, which specifies the mission, background, and output format. Then, the LLM agent is driven by the prompt requirement, and iteratively invokes WS-PSO-CM algorithm for exploration. Finally, the LLM agent autonomously terminates the loop and returns a set of hyper-parameters. Our experiment results show that the minimal sum-rate achieved by hyper-parameters generated via our LLM agent is significantly higher than those by both human heuristics and random generation methods. This indicates that an LLM agent with PSO knowledge and WS-PSO-CM algorithm background is useful in finding high-performance hyper-parameters.

Problem

• Automating hyper-parameter tuning for UAV communication algorithms
• Improving performance of WS-PSO-CM algorithm via LLM agent
• Reducing reliance on heuristic methods in parameter optimization

Innovation

• LLM agent automates hyper-parameter tuning
• Iterative framework with model context protocol
• Integrates WS-PSO-CM algorithm for exploration