GB-RVFL: Fusion of Randomized Neural Network and Granular Ball Computing

TL;DR

This paper introduces GB-RVFL, a novel neural network model that improves scalability and robustness by using granular balls instead of individual samples, and further enhances structure preservation with graph embedding.

Contribution

The paper proposes a new granular ball RVFL model and a graph embedding extension, addressing scalability, noise robustness, and geometric structure preservation in RVFL networks.

Findings

GB-RVFL outperforms baseline models on multiple datasets.

The model improves scalability by reducing matrix inversion complexity.

Graph embedding enhances topological structure preservation.

Abstract

The random vector functional link (RVFL) network is a prominent classification model with strong generalization ability. However, RVFL treats all samples uniformly, ignoring whether they are pure or noisy, and its scalability is limited due to the need for inverting the entire training matrix. To address these issues, we propose granular ball RVFL (GB-RVFL) model, which uses granular balls (GBs) as inputs instead of training samples. This approach enhances scalability by requiring only the inverse of the GB center matrix and improves robustness against noise and outliers through the coarse granularity of GBs. Furthermore, RVFL overlooks the dataset's geometric structure. To address this, we propose graph embedding GB-RVFL (GE-GB-RVFL) model, which fuses granular computing and graph embedding (GE) to preserve the topological structure of GBs. The proposed GB-RVFL and GE-GB-RVFL models are evaluated on KEEL, UCI, NDC and biomedical datasets, demonstrating superior performance compared to baseline models.