Convolutional Interval-Valued Neutrosophic Network for Intelligent Evaluation of Smart Clothing Design Choices

Abstract

Smart clothing design turned out to be a set of complex decision-making 
processes requiring complementary aesthetics, functionality, as well as customer 
preferences.  This, in turn, makes the traditional evaluation methods struggle to deal with 
uncertainty and subjectivity in customer feedback about smart clothing design.  In an 
attempt to address this challenge, this research article proposes a novel Neutrosophic 
approach that integrates Interval-Valued Neutrosophic (IVN) with Convolutional 
Network to build an intelligent tool for the evaluation of smart clothing design choices. 
The Neutrosophic representation enables modeling uncertainty, inconsistency, and 
hesitancy in decision-making by assigning interval-ed membership degrees for different 
views of smart clothes design. Using the interval-valued representations, we enable 
robust learning and interpretation of user partialities while handling vague feedback. 
Proof of concept experiments are conducted on a case study for a smart fashion dataset, 
and the quantitative results and analysis demonstrate that the proposed approach 
outperforms the standard techniques for smart clothes classification and ranking design 
choices. The findings from this analysis prove the ability of our approach to facilitate 
intelligent decision support in the fashion industry.

DOI: 10.5281/zenodo.14847504