Catastrophic Forgetting in Deep Graph Networks: an Introductory Benchmark for Graph Classification

In this work, we study the phenomenon of catastrophic forgetting in Building a robust machine learning model that incrementally learns the graph representation learning scenario. The primary objective from different tasks without forgetting requires methodologies of the analysis is to understand whether classical continual learning that account for drifts in the input distribution. The Continual techniques for flat and sequential data have a tangible impact on Learning (CL) research field addresses the catastrophic forgetting performances when applied to graph data. To do so, we experiment problem [15, 16] by devising learning algorithms that improve a with a structure-agnostic model and a deep graph network in a model's ability to retain previously gathered information. As of robust and controlled environment on three different datasets. The today, CL methods have been studied from the perspective of flat benchmark is complemented by an investigation on the effect of data [24, 28, 39] and, to a lesser extent, sequential data [11, 40].