Estimating Global Input Relevance and Enforcing Sparse Representations with a Scalable Spectral Neural Network Approach
This work addresses the need for explainable AI by providing a scalable method to estimate input relevance and enforce sparsity, which is incremental as it builds on spectral techniques for feature ranking.
The authors tackled the problem of identifying relevant input features in deep neural networks by proposing a spectral re-parametrization method that automatically ranks input importance during training, achieving sparse representations and improved explainability, with successful validation on synthetic and real data.
In machine learning practice it is often useful to identify relevant input features. Isolating key input elements, ranked according their respective degree of relevance, can help to elaborate on the process of decision making. Here, we propose a novel method to estimate the relative importance of the input components for a Deep Neural Network. This is achieved by leveraging on a spectral re-parametrization of the optimization process. Eigenvalues associated to input nodes provide in fact a robust proxy to gauge the relevance of the supplied entry features. Notably, the spectral features ranking is performed automatically, as a byproduct of the network training, with no additional processing to be carried out. Moreover, by leveraging on the regularization of the eigenvalues, it is possible to enforce solutions making use of a minimum subset of the input components, increasing the explainability of the model and providing sparse input representations. The technique is compared to the most common methods in the literature and is successfully challenged against both synthetic and real data.