Construction of Automated Machine Learning (AutoML) Framework Based on Large Language Models
Na Sun, Zidong Yu, Nan Jiang, Yuxiang Wang · 2025
As for automatic machine learning (AutoML), which can simplify the model design and adjustment process, continues to develop with the evolution of machine learning technology, and has become one of the factors promoting intelligent applications. Nevertheless, current AutoML frameworks still have a lot of limitations in many aspects when addressing more complex problems, particularly for settings where large-scale language models (LLMs) are utilized for automated learning. In this paper, we present a large-scale language model based automated machine learning framework. First of all, the framework integrates the existing automatic feature engineering and hyperparameter optimization technologies, and further utilizes the intelligent assistance of LLM model in the process of model generation, optimization and model inference to improve the accuracy and efficiency of the whole automation process. Its innovation is the deep fusion of large-scale language models and traditional AutoML working processes, and to automatically generate and fine-tune many machine learning models in multi-modal data and complex task scenarios based on their powerful contextual understanding and generation capabilities to realize more accurate and efficient modeling. Experimental results demonstrate that proposed framework improves the model's adaptive capacity and inference efficiency.