Predicting Fashion Sustainability Rate Using Hybrid Models and Sensitivity Analysis for Enhanced Accuracy and Feature Optimization
Year: 2026
Type: preprint
Abstract: Fashion sustainability promotes eco-friendly materials, waste reduction, fair labor, and low emissions. This study predicts sustainability rates using Machine Learning (ML) models and sensitivity analysis for improvement. Techniques, including the Stochastic Gradient Descent (SGD) and Logistic Regression Classification (LRC), benefit from advanced optimization methods like the Tunicate Swarm Optimization Algorithm (TSOA) and the Improved Arithmetic Optimization Algorithm (IAOA), which enhance accuracy and reliability. The goal is to identify the optimal model and key features. Sensitivity analysis is conducted on two models to identify the features that most influence performance. This allows for more informed decisions regarding feature selection and optimization. Through analysis, the SGIA model, achieving an accuracy score of 0.987 on the test phase and 0.983 overall, proves to be the most effective for predicting the fashion sustainability rate. Additionally, it appears that brand name and eco-friendly manufacturing practices are the most influential features in predicting sustainability.
Source: Authorea
Authors: Dongli WU, Lijing ZANG, Wenshuang LUO
Language: English
Subfield: Marketing
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7234917588
