Understanding the Role of Segmentation on Structure-Property Predictions Made via Machine Learning
Author(s)
Massey, Caroline Ellen
Advisor(s)
Editor(s)
Collections
Supplementary to:
Permanent Link
Abstract
Machine learning allows for the ability to predict an output from a diverse hyperspace of inputs. In the context of additive manufacturing, this class of approach could be useful in determining whether a specific measured defect field meets a given qualification requirement, this being particularly relevant for the aerospace and medical industries. The present study investigated the effect of porosity surface determination methods on performance of machine learning models used to predict the mechanical properties of AlSi10Mg processed by laser powder bed fusion from micro-computed tomography data. Machine learning models applied in this work include support vector machines, neural networks, decision trees, Bayesian classifiers, among others. The effects of isosurface thresholding and local gradient approaches for porosity segmentation, as well as image filtering schemes, on model precision was evaluated for samples produced under differing levels of global energy density (GED).
Sponsor
Date
2021-05-04
Extent
Resource Type
Text
Resource Subtype
Thesis