ASIA: an Autonomous System Identification Agent

2026-05-11

AI Summary

This work addresses the longstanding reliance on expert intuition in system identification for model selection, algorithm design, and hyperparameter tuning, where automated approaches have been lacking. It introduces ASIA, a novel framework that leverages large language model–driven autonomous agents to perform end-to-end system identification directly from natural language problem descriptions, encompassing hypothesis generation, code implementation, and performance evaluation in a closed loop. Experimental results on two standard benchmarks demonstrate that ASIA effectively discovers high-quality dynamic models and training strategies, thereby validating its automation capability. The study also uncovers critical challenges such as test leakage and transparency, highlighting important directions for future research in autonomous system identification.

Abstract

Over the years, research in system identification has provided a rich set of methods for learning dynamical models, together with well-established theoretical guarantees. In practice, however, the choice of model class, training algorithm, and hyperparameter tuning is still largely left to empirical trial-and-error, requiring substantial expert time and domain experience. Motivated by recent advances in agentic artificial intelligence, we present ASIA, a framework that delegates this iterative search to a large language model acting as an autonomous coding agent. Building on existing agentic platforms, ASIA closes the loop between hypothesis, implementation, and evaluation without human intervention, requiring only a plain-English description of the identification problem. We conduct an empirical study of ASIA on two system identification benchmarks and analyse the agent's search behaviour, the architectures and training strategies it discovers, and the quality of the resulting models. We also discuss the potential of the approach and its current limitations, including implicit test leakage, reduced methodological transparency, and reproducibility concerns.

Problem

• system identification
• model selection
• hyperparameter tuning
• autonomous agent
• trial-and-error

Innovation

• autonomous agent
• system identification
• large language model
• automated machine learning
• agentic AI