A Data-Centric Perspective on Evaluating Machine Learning Models for Tabular Data

TL;DR

This paper advocates for a data-centric evaluation approach in machine learning for tabular data, emphasizing dataset-specific preprocessing, feature engineering, and temporal adaptation, which significantly influence model performance and rankings.

Contribution

It introduces a data-centric evaluation framework with expert-level preprocessing pipelines, highlighting the impact of data handling on model assessment and ranking stability.

Findings

Feature engineering alters model rankings and reduces performance gaps.

Manual feature engineering benefits both tree-based models and neural networks.

Adapting to distribution shifts is important even in static tabular data.

Abstract

Tabular data is prevalent in real-world machine learning applications, and new models for supervised learning of tabular data are frequently proposed. Comparative studies assessing the performance of models typically consist of model-centric evaluation setups with overly standardized data preprocessing. This paper demonstrates that such model-centric evaluations are biased, as real-world modeling pipelines often require dataset-specific preprocessing and feature engineering. Therefore, we propose a data-centric evaluation framework. We select 10 relevant datasets from Kaggle competitions and implement expert-level preprocessing pipelines for each dataset. We conduct experiments with different preprocessing pipelines and hyperparameter optimization (HPO) regimes to quantify the impact of model selection, HPO, feature engineering, and test-time adaptation. Our main findings are: 1. After dataset-specific feature engineering, model rankings change considerably, performance differences decrease, and the importance of model selection reduces. 2. Recent models, despite their measurable progress, still significantly benefit from manual feature engineering. This holds true for both tree-based models and neural networks. 3. While tabular data is typically considered static, samples are often collected over time, and adapting to distribution shifts can be important even in supposedly static data. These insights suggest that research efforts should be directed toward a data-centric perspective, acknowledging that tabular data requires feature engineering and often exhibits temporal characteristics. Our framework is available under: https://github.com/atschalz/dc_tabeval.