CORTEXA
← Browse
openalexOpen MIND2026-07-23

Utilizing Concept Ontologies for Designing Neural Network-Based Classifiers

Kamil Szwed

Deep neural networks have achieved substantial success in image, text, and signal analysis, but their advantage is less consistent for heterogeneous tabular data, where tree-based ensemble methods often remain strong baselines. This study proposes CANON (Cross-Attention Neuro-symbolic Ontology Network), a neuro-symbolic architecture that integrates a hierarchical a priori concept ontology with a modular mixture-of-experts mechanism. CANON is designed to combine data-driven representation learning with explicit domain structure and to reduce the influence of irrelevant or weakly informative features. The architecture was evaluated on a comprehensive dataset of medical records and compared with selected tree-based machine learning methods and deep neural network models. The experimental results show that CANON achieved the best predictive performance among the evaluated methods while maintaining an explicit correspondence between groups of clinical concepts and individual model components. These findings indicate that ontology-guided neural architectures can provide an effective basis for accurate and interpretable clinical decision support and may also be useful in other regulated domains in which predictive models must remain consistent with domain-specific knowledge.

Related papers

openalexOpen MIND2026-07-23

The Volatility Channel (Financial Branch)

Jean-Pierre Bronsard

🔗 Reproducible code: github.com/jpbronsard/syntonic-portfolio v 3.0 V2.0 measured one channel of financial adaptation: the return channel, where \(\tau^\star=1/\sqrt{2}\) is the structural signature of the random-walk limit. Markets have a second channel with the same structure…