Novel Representation Learning Technique Using Graphs for Performance Analytics
Published in 2023 International Conference on Machine Learning and Applications (ICMLA), 2023
Download here
A novel graph-based representation learning approach that transforms tabular HPC performance data into graphs, enabling Graph Neural Network (GNN) techniques to capture complex relationships between features and samples. Addresses regression problems in performance analytics, such as predicting application execution time.
Authors: Tarek Ramadan, Ankur Lahiry, and Tanzima Z Islam