Revisiting Deep Learning Models for Tabular Data
This work addresses the problem of inconsistent benchmarking for researchers and practitioners in tabular data analysis, but it is incremental as it builds on existing architectures.
The authors tackled the lack of consistent comparisons and effective baselines in deep learning for tabular data by identifying two simple architectures—a ResNet-like model and a Transformer adaptation—that outperform many existing methods under standardized protocols, though no universally superior solution was found compared to Gradient Boosted Decision Trees.
The existing literature on deep learning for tabular data proposes a wide range of novel architectures and reports competitive results on various datasets. However, the proposed models are usually not properly compared to each other and existing works often use different benchmarks and experiment protocols. As a result, it is unclear for both researchers and practitioners what models perform best. Additionally, the field still lacks effective baselines, that is, the easy-to-use models that provide competitive performance across different problems. In this work, we perform an overview of the main families of DL architectures for tabular data and raise the bar of baselines in tabular DL by identifying two simple and powerful deep architectures. The first one is a ResNet-like architecture which turns out to be a strong baseline that is often missing in prior works. The second model is our simple adaptation of the Transformer architecture for tabular data, which outperforms other solutions on most tasks. Both models are compared to many existing architectures on a diverse set of tasks under the same training and tuning protocols. We also compare the best DL models with Gradient Boosted Decision Trees and conclude that there is still no universally superior solution.