A Research on Visualization Method of Middle-aged Women's Wear User Demand Data Based on t-SNE

Abstract

The high-dimensional, fuzzy and sparse characteristics of clothing users often bring inconvenience to designers. With this regard, this paper will attempt to improve the rationality and accuracy of the analysis method of the clothing users’ demand by applying computers technology of t-distribution stochastic neighbor embedding (t-SNE) in order to solve the problem of analysis and processing of high-dimensional data by reducing the dimension of data. Starting from the middle-aged women's wearing users in China, this paper uses t-SNE to map the complex and massive demand information relationship among users to 3D space, and combines the visualization method to show the closeness and relevance of the demand information. The correlation between the needs of middle-aged women's wear users in China and their age, disposable income, purchase frequency is also studied in this research together with the evaluation of the relevance so as to verify the effectiveness of the method. This paper proposes a visualization method to analyze the needs of clothing users, getting rid of the reliance on "experience ", from human subjective judgment to machine recognition, and provides a reference for the research of perfecting the needs of middle-aged women's wear users.