A Study On The Design Of Garment Parts Based On Multiple Linear Regression

With the changing of information technology,the era of big data means that all kinds of information can be integrated into a huge container for statistical analysis,rapid classification queries and calls.At the same time,the continuous updating of information technology and increasing cross-disciplines make clothing popular prediction mechanism have to get rid of the traditional method of artificial finishing,subjective prediction,to explore a new way out.For the clothing market,the relevant research on clothing parts design is still slightly inadequate,such as the complete system of parts design database,computer technology in clothing parts trend prediction application and component design intelligent trend prediction system,etc.have yet to be supplemented and perfected;In the face of accelerating trend changes,the need for clothing research and development design,the field of clothing urgently needs efficient and intelligent tools to predict trends in science and specialty.Using mathematical statistics and computer science methods,on the basis of collecting 12,662 runway pictures of international well-known women's clothing brands in the past five years,this topic uses computer technology to build a database of parts,analyzes and preprocesses the data with the help of Python library functions,uses sklearn algorithm to establish a multilinear regression model,uses machine learning to train models,then get the multi-regression equation and relatively popular prediction values and fitting line chart,finally get the popular law and design tendency within five years.At last,complete create inspiration,the trend analysis of color,fabric,detail and the whole series of graduation design works.