# CA-HFP: Curvature-Aware Heterogeneous Federated Pruning with Model Reconstruction Source: [CA-HFP: Curvature-Aware Heterogeneous Federated Pruning with Model Reconstruction]() Collection index: [llms.txt]() Data updated: 2026-03-17T05:00:04.157Z Authors: Gang Hu, Yinglei Teng, Pengfei Wu, Shijun Ma Published: 2026-03-13T02:45:00.000Z ## Abstract Federated learning on heterogeneous edge devices requires personalized compression while preserving aggregation compatibility and stable convergence. We present Curvature-Aware Heterogeneous Federated Pruning (CA-HFP), a practical framework that enables each client perform structured, device-specific pruning guided by a curvature-informed significance score, and subsequently maps its compact submodel back into a common global parameter space via a lightweight reconstruction. We derive a convergence bound for federated optimization with multiple local SGD steps that explicitly accounts for local computation, data heterogeneity, and pruning-induced perturbations; from which a principled loss-based pruning criterion is derived. Extensive experiments on FMNIST, CIFAR-10, and CIFAR-100 using VGG and ResNet architectures under varying degrees of data heterogeneity demonstrate that CA-HFP preserves model accuracy while significantly reducing per-client computation and communication costs, outperforming standard federated training and existing pruning-based baselines. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Image Classification on FMNIST (test) with Alpha-Weighted Accuracy](): Image Classification; fMNIST (test); Accuracy (alpha=0.1): 89. - [Image Classification on CIFAR100 (test) (Alpha-Weighted Accuracy)](): Image Classification; CIFAR100 (test); Accuracy (alpha=0.1): 56.27. - [Image Classification on CIFAR10 (test) (Alpha-Weighted Accuracy)](): Image Classification; CIFAR10 (test); Accuracy (alpha=0.1): 73.12. ## Original sources and implementation [Original paper]() [PDF]()