Locally Sparse Neural Networks for Tabular Biomedical Data
This addresses the need for interpretable and robust models in biomedicine, where practitioners often avoid neural networks due to overfitting and opacity, though it is incremental as it builds on existing sparse and interpretable methods.
The paper tackled the problem of neural networks overfitting and lacking interpretability on tabular biomedical data by proposing a locally sparse neural network that selects sample-specific features, resulting in outperforming state-of-the-art models in performance and interpretability on synthetic and real-world datasets.
Tabular datasets with low-sample-size or many variables are prevalent in biomedicine. Practitioners in this domain prefer linear or tree-based models over neural networks since the latter are harder to interpret and tend to overfit when applied to tabular datasets. To address these neural networks' shortcomings, we propose an intrinsically interpretable network for heterogeneous biomedical data. We design a locally sparse neural network where the local sparsity is learned to identify the subset of most relevant features for each sample. This sample-specific sparsity is predicted via a \textit{gating} network, which is trained in tandem with the \textit{prediction} network. By forcing the model to select a subset of the most informative features for each sample, we reduce model overfitting in low-sample-size data and obtain an interpretable model. We demonstrate that our method outperforms state-of-the-art models when applied to synthetic or real-world biomedical datasets using extensive experiments. Furthermore, the proposed framework dramatically outperforms existing schemes when evaluating its interpretability capabilities. Finally, we demonstrate the applicability of our model to two important biomedical tasks: survival analysis and marker gene identification.