Neural Network Implementation of the Renormalization Group for Fault Diagnosis with Class Imbalance
arXiv:2606.18326v1 Announce Type: new Abstract: The application of machine learning models in practical tasks faces challenges such as class imbalance and multidimensional noise. This paper proposes RGNet, a neural network architecture based on the concept of the renormalization group (RG), for hierarchical coarse-graining of the feature space. The model sequentially compresses the input dimensionality and concatenates all scales before classification, allowing it to capture both local details and global patterns. The notion of RG-flows is introduced - interpretable low-dimensional representat
The continuous evolution of AI and machine learning faces persistent challenges with data quality and class imbalance, driving research into novel architectural solutions like those inspired by the renormalization group.
This research introduces a new neural network architecture, RGNet, that promises more robust fault diagnosis even with imperfect data, potentially improving reliability and efficiency in various industrial and scientific applications.
The approach of hierarchical coarse-graining and concatenating features before classification offers a new method to capture both local and global data patterns, enhancing interpretability and performance in complex systems.
- · AI researchers
- · Manufacturing industries
- · Healthcare diagnostics
- · Predictive maintenance sector
- · Traditional fault diagnosis methods
- · Systems highly sensitive to class imbalance
Improved accuracy and reliability in machine learning models for fault diagnosis across various domains.
Reduced operational downtime and maintenance costs in industrial settings due to more effective anomaly detection.
Accelerated adoption of AI in critical infrastructure where data quality and imbalance are significant hurdles.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG