Local Learning with Boosting-based Backpropagation-Free Graph Neural Networks

The framework of Backpropagation-Free Graph Neural Networks (BF-GNNs) enables local learning at the neuron level in GNNs. While BF-GNNs can match the performance of their backpropagation-based counterparts, they may develop redundant internal representations that limit further gains. To address this issue, we propose an innovative architecture dubbed Boosting-based Backpropagation-Free GNN (B3F-GNN), where each network module contains multiple backpropagation-free neurons trained locally and combined as a classifier. Within each layer, later modules exploit error signals from earlier trained modules to refine predictions by diversifying internal representations. We implement this approach with two complementary boosting strategies: sample reweighting, in the spirit of AdaBoost, and error-guided prototype selection for gating, which concentrates non-linearities where previous modules struggled. The modular design also enables any-time incremental training by adding more modules on demand within resource constraints. An ablation study and in-depth experimental analysis show that both strategies reduce redundancy and increase specialization, leading to statistically significant accuracy improvements over backpropagation-based counterparts on standard node-classification benchmarks.