*Online data repositories provide designers and engineers with a convenient source of data. Over time, the wealth and type of readily-available data within online repositories has greatly expanded. This data increase permits new uses for machine learning approaches which rely on large, high-dimensional datasets. However, a comparison of the efficacies of attribute-based data, which lends itself well to traditional machine learning algorithms, and form-based data, which lends itself to powerful deep machine learning algorithms, has not fully been established. This paper presents one such comparison for an exemplar dataset. As the efficacy of different machine learning approaches may be dependent on the specific application, this method is intended to lay the groundwork for future studies that produce more extensive comparisons. Specifically, this work makes use of a manufactured gear dataset for sale price prediction. Two traditional machine learning algorithms, M5Rules and SMOreg, are selected due to their applicability to the gear attribute-based data. These algorithms are compared to a neural network model that is trained on a voxelized version of the gears’ 3D models, the form-based data. Results show that both data types provide comparable predictive accuracy.*


Comparing Attribute- and Form-Based Machine Learning Techniques for Component Prediction

Division / Sponsors: 
Design Engineering Division and Computers & Information In Engineering Division


Paper Proceedings: https://asmedigitalcollection.asme.org/IDETC-CIE/IDETC-CIE2020/volume/83907

IDETC2020 - Virtual Conference

Supporting the need among researchers worldwide for additional online and multimedia content, ASME now offers the ASME Conference Video Collection. This platform will collect and provide long-term access to conference presentation videos, encompassing all technology areas within ASME's technical conference and events program. About 30,000-40,000 engineers across all mechanical engineering and related disciplines participate in ASME conferences each year. This includes authors of approximately 6,000 to 8,000 technical papers published by ASME, the majority of whom also submit video presentations for our virtual conferences. The video collection will be available as an add-on to existing product packages or as a separate subscription.