Graph Neural Networks in TensorFlow and Keras with Spektral
arXiv:2006.12138
Abstract
In this paper we present Spektral, an open-source Python library for building graph neural networks with TensorFlow and the Keras application programming interface. Spektral implements a large set of methods for deep learning on graphs, including message-passing and pooling operators, as well as utilities for processing graphs and loading popular benchmark datasets. The purpose of this library is to provide the essential building blocks for creating graph neural networks, focusing on the guiding principles of user-friendliness and quick prototyping on which Keras is based. Spektral is, therefore, suitable for absolute beginners and expert deep learning practitioners alike. In this work, we present an overview of Spektral's features and report the performance of the methods implemented by the library in scenarios of node classification, graph classification, and graph regression.
arXiv commentICML 2020 - GRL+ Workshop
Cited by in corpus (9)
- Graph Neural Networks with convolutional ARMA filters
- Hierarchical Representation Learning in Graph Neural Networks with Node Decimation Pooling
- Vertex and Energy Reconstruction in JUNO with Machine Learning Methods
- Implementing graph neural networks with TensorFlow-Keras
- IGNNITION: Bridging the Gap Between Graph Neural Networks and Networking Systems
- Image2Flow: A hybrid image and graph convolutional neural network for rapid patient-specific pulmonary artery segmentation and CFD flow field calculation from 3D cardiac MRI data
- gSuite: A Flexible and Framework Independent Benchmark Suite for Graph Neural Network Inference on GPUs
- A Method to Predict Semantic Relations on Artificial Intelligence Papers
- Classification of vertices on social networks by multiple approaches