An efficient way of material classification using uncertainty assessment

Abstract

Analysis of material properties using the available data is a compulsory part of the material classification and is considered as a challenging task in industrial processes. Classification refers to the segregation of objects using known data or knowledge. The uncertainties in the data increase the complexities of material classification. In this paper, we present a methodology to measure uncertainties in the attributes-values. An example illustrates the efficiency, validity, and effectiveness of the proposed method. We find the most significant features for the classification of glass materials using uncertainty evaluation and optimization.