Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Peng, Furong, Gao, Jinzhen, Lu, Xuan, Liu, Kang, Huo, Yifan, Wang, Sheng
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2506.17576
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913953722925056
author Peng, Furong
Gao, Jinzhen
Lu, Xuan
Liu, Kang
Huo, Yifan
Wang, Sheng
author_facet Peng, Furong
Gao, Jinzhen
Lu, Xuan
Liu, Kang
Huo, Yifan
Wang, Sheng
contents Graph Convolutional Networks (GCNs) suffer from severe performance degradation in deep architectures due to over-smoothing. While existing studies primarily attribute the over-smoothing to repeated applications of graph Laplacian operators, our empirical analysis reveals a critical yet overlooked factor: trainable linear transformations in GCNs significantly exacerbate feature collapse, even at moderate depths (e.g., 8 layers). In contrast, Simplified Graph Convolution (SGC), which removes these transformations, maintains stable feature diversity up to 32 layers, highlighting linear transformations' dual role in facilitating expressive power and inducing over-smoothing. However, completely removing linear transformations weakens the model's expressive capacity. To address this trade-off, we propose Layer-wise Gradual Training (LGT), a novel training strategy that progressively builds deep GCNs while preserving their expressiveness. LGT integrates three complementary components: (1) layer-wise training to stabilize optimization from shallow to deep layers, (2) low-rank adaptation to fine-tune shallow layers and accelerate training, and (3) identity initialization to ensure smooth integration of new layers and accelerate convergence. Extensive experiments on benchmark datasets demonstrate that LGT achieves state-of-the-art performance on vanilla GCN, significantly improving accuracy even in 32-layer settings. Moreover, as a training method, LGT can be seamlessly combined with existing methods such as PairNorm and ContraNorm, further enhancing their performance in deeper networks. LGT offers a general, architecture-agnostic training framework for scalable deep GCNs. The code is available at [https://github.com/jfklasdfj/LGT_GCN].
format Preprint
id arxiv_https___arxiv_org_abs_2506_17576
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Towards a deeper GCN: Alleviate over-smoothing with iterative training and fine-tuning
Peng, Furong
Gao, Jinzhen
Lu, Xuan
Liu, Kang
Huo, Yifan
Wang, Sheng
Machine Learning
Graph Convolutional Networks (GCNs) suffer from severe performance degradation in deep architectures due to over-smoothing. While existing studies primarily attribute the over-smoothing to repeated applications of graph Laplacian operators, our empirical analysis reveals a critical yet overlooked factor: trainable linear transformations in GCNs significantly exacerbate feature collapse, even at moderate depths (e.g., 8 layers). In contrast, Simplified Graph Convolution (SGC), which removes these transformations, maintains stable feature diversity up to 32 layers, highlighting linear transformations' dual role in facilitating expressive power and inducing over-smoothing. However, completely removing linear transformations weakens the model's expressive capacity. To address this trade-off, we propose Layer-wise Gradual Training (LGT), a novel training strategy that progressively builds deep GCNs while preserving their expressiveness. LGT integrates three complementary components: (1) layer-wise training to stabilize optimization from shallow to deep layers, (2) low-rank adaptation to fine-tune shallow layers and accelerate training, and (3) identity initialization to ensure smooth integration of new layers and accelerate convergence. Extensive experiments on benchmark datasets demonstrate that LGT achieves state-of-the-art performance on vanilla GCN, significantly improving accuracy even in 32-layer settings. Moreover, as a training method, LGT can be seamlessly combined with existing methods such as PairNorm and ContraNorm, further enhancing their performance in deeper networks. LGT offers a general, architecture-agnostic training framework for scalable deep GCNs. The code is available at [https://github.com/jfklasdfj/LGT_GCN].
title Towards a deeper GCN: Alleviate over-smoothing with iterative training and fine-tuning
topic Machine Learning
url https://arxiv.org/abs/2506.17576