---
title: Edge Directionality Improves Learning on Heterophilic Graphs
url: https://www.ml-quant.com/papers/arxiv/2305.10498/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-09-26
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
identifier: arXiv:2305.10498
source_url: https://arxiv.org/abs/2305.10498
featured: 2023-12-06
citations: 148
topic: ML & AI Methods
---


# Edge Directionality Improves Learning on Heterophilic Graphs

The study presents Directed Graph Neural Network (Dir-GNN), a new deep learning framework for directed graphs that surpasses traditional models in heterophilic benchmarks.

- Source: https://arxiv.org/abs/2305.10498
- Identifier: arXiv:2305.10498
- Released: 2023-05-17
- First featured: Quant Letter No. 28 (2023-12-06): https://www.ml-quant.com/issues/2023-12-06/
- Citations (Semantic Scholar): 148
- Published in: LOG IN
- Topic: ML & AI Methods

## Related

- [GraphCNNpred: A stock market indices prediction using a Graph based deep learning system](https://www.ml-quant.com/papers/arxiv/2407.03760/): Stock Prediction: The paper introduces a graph neural network-based convolutional neural network model for predicting stock market prices, using custom feature engineering on diverse data sources.
- [Graph Mamba: Towards Learning on Graphs with State Space Models](https://www.ml-quant.com/papers/arxiv/2402.08678/): Graph Mamba Networks, a new type of Graph Neural Networks, have been introduced, which achieve excellent performance in various benchmark datasets despite lower computational cost.
- [Modular Duality in Deep Learning](https://www.ml-quant.com/papers/arxiv/2410.21265/): The article presents a new theory of modular dualization for general neural networks, providing a theoretical basis for fast and scalable training algorithms, potentially leading to a new generation of optimizers for neural architectures.
- [Historical calibration of SVJD models with deep learning](https://www.ml-quant.com/papers/ssrn/4650097/): The paper suggests using deep neural networks to calibrate parameters of Stochastic Volatility Jump Diffusion models, proving to be more accurate, robust, and faster than other methods.
- [Can overfitted deep neural networks in adversarial training generalize? - An approximation viewpoint](https://www.ml-quant.com/papers/arxiv/2401.13624/): The study offers a theoretical insight into the robust overfitting issue in adversarial training on over-parameterized deep neural networks (DNNs), showing that overfitting can be avoided and a robust generalization gap is unavoidable, with the model capacity requirement depending on the target function's smoothness.
- [GNN-VPA: A Variance-Preserving Aggregation Strategy for Graph Neural Networks](https://www.ml-quant.com/papers/arxiv/2403.04747/): A new variance-preserving aggregation function for Graph Neural Networks is proposed, which could lead to more efficient and self-normalizing networks.