The Research And Applications About Subsidiary Decision Of Productive Apparel Sizes And Styles Basing On Technology Of Data Mining

Under the current market environment accompanying increasingly intense competition, it is technique focal point to the existence and development of clothing enterprise to establish scientific and reasonable apparel sizes and styles form,it has the great significance not only for satisfying the consumer request, but also towards raising the economic performance of enterprise. The clothing enterprises have own apparel sizes and styles form generally, the quantity also is different during clothing made. However the clothing factory's common goal is adapting more figures of consumers but not increase cost or increase cost a little, it's meaning is establishing a series of apparel sizes and styles of " fewer apparel sizes and styles number, more overlay proportion", then lower production cost, and enhance production efficiency, it is also the purpose that this topic wants to study.This topic applies cluster analysis method of data mining, bases on a great deal of figure measure data that accumulated by enterprise long-term, discovers the proportion rules of the apparel sizes and styles and the quantities of crowd figure, then guides to establish a scientific and reasonable apparel sizes and styles form; Through analyzing historical sale data and applying time-series data mining algorithm, carries on the prediction of costume sales volume, discovers the hidden rule and modes, then provides the basis to the production decision of apparel sizes and styles, and combining the overlay proportion of crowd figure, predicts the future trend to produce further.This text takes the research and applications of data mining algorithm as the core (the emphasis is cluster and time-series algorithm), takes the design and realizations of" Subsidiary Decision System of Productive Apparel Sizes and Styles " as the main line, researches to upbuild an subsidiary decision model of apparel sizes and styles. This research obtained the following result:1. Establish a mathematics decision model of the apparel sizes and stylesThis project tries to establish the relation model between figure measure data cluster and apparel sizes and styles basing on the vast volume data warehouse by the data mining cluster algorithm, brings to bear vendition data resources, restricts together the distribution of apparel sizes and styles and yield for correspond of it during the process of production.2. Provid the solution and softwares realization platform for establishinga regional enterprise apparel sizes and styles systemThis research clusters effectively according to regional figure measure data warehouse, it provide softwares realization platform for suiting the costume enterprise oneself characteristics, forming a system of the productive apparel sizes and styles of the particular style, instituting the costume custom-tailoring and commercial productive plan to fit in with regional crowd figure.3.Provide the technique support for subsidiary decision of productive apparel sizes and stylesBasing on the reacher of mathematics decision model of the apparel sizes and styles, realize the analysis of basic proportion relation between apparel sizes and styles and crowd amount, using sales data resources and the product life cycle theories, basing on Time-series algorithm, establish product life cycle model and the sales forecasting model, control the distribution of apparel sizes and styles during produce and yield of each number, aim at the concrete clothing style carries on the decision guiding at apparel sizes and styles and quantity to produce.