Performance Evaluation of Predictive Classifiers For Knowledge Discovery From Engineering Materials Data Sets

2012/09/12 by Hemanth K. S Doreswamy, Doreswamy, Hemanth K. S
Computer Science · Engineering · #FOS: Computer and information sciences #Fault Detection and Control Systems #Machine Learning (cs.LG) #Neural Networks and Applications #Rough Sets and Fuzzy Logic #cs.LG

paper · pdf · doi:10.48550/arxiv.1209.2501

Volume 3,No 3,March 2011 7 pages, 6 tables, 4 figures

arxiv created 2012/09/12 · openalex publication_date 2012/09/12 · arxiv updated 2012/09/20 · openalex created_date 2016/06/24 · openalex updated_date 2026/07/28

Abstract

In this paper, naive Bayesian and C4.5 Decision Tree Classifiers(DTC) are successively applied on materials informatics to classify the engineering materials into different classes for the selection of materials that suit the input design specifications. Here, the classifiers are analyzed individually and their performance evaluation is analyzed with confusion matrix predictive parameters and standard measures, the classification results are analyzed on different class of materials. Comparison of classifiers has found that naive Bayesian classifier is more accurate and better than the C4.5 DTC. The knowledge discovered by the naive bayesian classifier can be employed for decision making in materials selection in manufacturing industries.

Related

• Hybrid Data Mining Technique for Knowledge Discovery from Engineering Materials' Data sets
• A review of the application of machine learning and data mining approaches in continuum materials mechanics
• Machine learning application in the life time of materials
• Leveraging Composition-Based Material Descriptors for Machine Learning Optimization
• Position Paper on Materials Design -- A Modern Approach