Large Language Models versus Classical Machine Learning: Performance in COVID-19 Mortality Prediction Using High-Dimensional Tabular Data

TL;DR

This study compares classical machine learning models and large language models in predicting COVID-19 mortality from high-dimensional tabular data, highlighting the strengths and limitations of each approach.

Contribution

It demonstrates that fine-tuning LLMs can significantly improve their performance, but classical models still outperform LLMs on structured data tasks.

Findings

XGBoost and RF achieved F1 scores of 0.87 and 0.83.

GPT-4 achieved an F1 score of 0.43 in zero-shot classification.

Fine-tuning Mistral-7b increased recall from 1% to 79%, with a stable F1 of 0.74.

Abstract

This study compared the performance of classical feature-based machine learning models (CMLs) and large language models (LLMs) in predicting COVID-19 mortality using high-dimensional tabular data from 9,134 patients across four hospitals. Seven CML models, including XGBoost and random forest (RF), were evaluated alongside eight LLMs, such as GPT-4 and Mistral-7b, which performed zero-shot classification on text-converted structured data. Additionally, Mistral- 7b was fine-tuned using the QLoRA approach. XGBoost and RF demonstrated superior performance among CMLs, achieving F1 scores of 0.87 and 0.83 for internal and external validation, respectively. GPT-4 led the LLM category with an F1 score of 0.43, while fine-tuning Mistral-7b significantly improved its recall from 1% to 79%, yielding a stable F1 score of 0.74 during external validation. Although LLMs showed moderate performance in zero-shot classification, fine-tuning substantially enhanced their effectiveness, potentially bridging the gap with CML models. However, CMLs still outperformed LLMs in handling high-dimensional tabular data tasks. This study highlights the potential of both CMLs and fine-tuned LLMs in medical predictive modeling, while emphasizing the current superiority of CMLs for structured data analysis.