Intelligent Classification of Pure and Nano-Composite PMMA Based on Thermal Properties

PMMA has become a paramount polymer with many advanced applications. Though, no thermal NDT method is defined for screening this significant polymer from its composites. Therefore, in this survey, we planned an experimental test coupled with artificial intelligence to classify the specimens into pure and composite classes. The test is quick and financially acceptable. More than ۳۰۰۰ neural pattern recognition networks with ۱۱ distinct architectures were tested and an unbalanced data effect removal technique was utilized. The results showed that the weighted network with ۳ neurons could successfully classify the samples with ۸۵% accuracy.

کلیدواژه ها:

Machine Learning ، Simulation ، PMMA bone cement ، NDT method ، Fault Classification.