Assessing Machine Learnability of Image and Graph Representations for Drone Performance Prediction

DS 116: Proceedings of the DESIGN2022 17th International Design Conference

Year: 2022
Editor: Mario Štorga, Stanko Škec, Tomislav Martinec, Dorian Marjanović
Author: Binyang Song (1), Christopher Mccomb (2), Faez Ahmed (1)
Series: DESIGN
Institution: 1: Massachusetts Institute of Technology, United States of America; 2: Carnegie Mellon University, United States of America
Section: Artificial Intelligence and Data-Driven Design
Page(s): 1777-1786
DOI number: https://doi.org/10.1017/pds.2022.180
ISSN: 2732-527X (Online)

Abstract

Deep learning (DL) from various representations have succeeded in many fields. However, we know little about the machine learnability of distinct design representations when using DL to predict design performance. This paper proposes a graph representation for designs and compares it to the common image representation. We employ graph neural networks (GNNs) and convolutional neural networks (CNNs) respectively to learn them to predict drone performance. GCNs outperform CNNs by 2.6-8.1% in predictive validity. We argue that graph learning is a powerful and generalizable method for such tasks.

Keywords: engineering design, artificial intelligence (AI), design evaluation, design representation, graph convolutional networks