A COMPREHENSIVE EXPLORATION THROUGH IG GAIN COEFFICIENT CALCULATION AND COSINE SIMILARITY CALCULATION: SYMBOLIC FEATURE ANALYSIS OF BIONICS FURNITURE DESIGN

• Published in: Journal of Nonlinear and Convex Analysis
• Main Author: Lin X.
• Format: Article
• Language: English
• Published: Yokohama Publications 2024
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85196361277&partnerID=40&md5=cab7c14081568b64be3174bfb044d3b4

This study introduces a comprehensive symbolic feature analysis method for bionics furniture design, utilising IG gain coefficient and cosine similarity calculations. Initially, the IG gain coefficient is computed to assess the contribution levels of various symbolic features to the design objectives. Based on these evaluations, the most significant symbol features in the design are identified. Subsequently, cosine similarity calculation is employed to analyse the selected symbolic features in the design case. This holistic approach enables designers to systematically choose and optimise symbolic features, enhancing the scien-tificity and consistency of the design process. Experimental results demonstrate the efficacy of the proposed method in enhancing data-driven design decisions. Furthermore, the method offers designers innovative and analogous approaches to symbolic feature analysis, presenting new ideas and methodologies for the advancement of bionics furniture design. The comprehensive nature of this calculation method provides a practical framework for designers to make informed decisions in symbolic feature analysis, promising notable outcomes in practical design applications. © 2024 Yokohama Publications. All rights reserved.