Explicit neural network classifiers for non-separable data
arXiv:2504.18710v21 citationsh-index: 3
Originality Synthesis-oriented
AI Analysis
This addresses classification challenges for non-separable datasets, but appears incremental as it builds on existing network characterizations.
The paper tackles the problem of classifying non-separable data by characterizing feedforward neural networks with truncation maps, and shows that a ReLU network can implement a feature map to separate concentric data.
We fully characterize a large class of feedforward neural networks in terms of truncation maps. As an application, we show how a ReLU neural network can implement a feature map which separates concentric data.