We present a method to integrate Large Language Models (LLMs) and traditional tabular data classification techniques, addressing LLMs challenges like data serialization sensitivity and biases. We introduce two strategies utilizing LLMs for ranking categorical variables and generating priors on correlations between continuous variables and targets, enhancing performance in few-shot scenarios. We focus on Logistic Regression, introducing MonotonicLR that employs a non-linear monotonic function for mapping ordinals to cardinals while preserving LLM-determined orders. Validation against baseline models reveals the superior performance of our approach, especially in low-data scenarios, while remaining interpretable.
@article{arxiv.2311.11628,
title = {Incorporating LLM Priors into Tabular Learners},
author = {Max Zhu and Siniša Stanivuk and Andrija Petrovic and Mladen Nikolic and Pietro Lio},
journal= {arXiv preprint arXiv:2311.11628},
year = {2023}
}
Comments
Table Representation Learning Workshop at NeurIPS 2023