基于非参数核估计模型的服装销售预测研究

发布时间：2019-03-01 08:23

【摘要】：服装市场是典型的供远大于需的买方市场,在激烈的市场竞争中,属于短生命周期产品的服装,具有较长提前期、较短销售期、较低期末残值、快速需求变化等特点,使得服装供应商难以准确预测市场需求和销量走势,我国传统的订货会模式也无法快速响应服装流行趋势和消费者喜好的变换。很多服装企业在进行市场预测时,往往缺乏科学周密的计划方案,凭借个人经验和主观想法作出定性判断,预测准确性低,盲目地订货和补货使得产量与未来销售状况不匹配,可能产生缺货风险。然而更多的情况是服装企业将面临库存积压,增加了运营成本,减少了利润收入,影响自身的长远发展。本文的研究对象是国内某家一线品牌的时尚休闲服装公司,生命周期短的特点在时尚类服装上显得更为突出。本文结合该服装公司的产品特征和销售数据,通过数学方法来解决其销量预测方面的难题,寻找可靠的预测方案提高时尚服装销售的预测精度。本文主要的研究成果,是运用MATLAB作为编程工具,为服装公司产品的生命周期进行阶段划分,以及设计非参数核密度估计模型进行预测。利用非参数核估计函数预测该公司各个品类服装进入生命期后的日销量和总销量,并对预测结果的准确性进行了检验。在结果有效的基础上,根据预测值对公司订货决策提出建议。本文的研究可以帮助服装公司建立和完善预测系统,为企业合理安排生产补货计划提供参考。
[Abstract]:Clothing market is a typical buyer's market. In the fierce market competition, clothing belongs to short life cycle products. It has the characteristics of longer lead time, shorter sales period, lower end-of-life residual value, rapid demand change, and so on, and the clothing market has the characteristics of long lead time, shorter sales period, lower end-of-term residual value and rapid demand change. It makes it difficult for clothing suppliers to accurately predict the market demand and sales trend, and the traditional order meeting pattern in China can not respond quickly to the change of fashion trend and consumer preference. When making market forecast, many garment enterprises often lack scientific and careful planning scheme, make qualitative judgment by virtue of personal experience and subjective thought, forecast accuracy is low, blindly order and replenish make output and future sales situation not match, There may be a risk of out-of-stock. However, more clothing companies will face a backlog of inventory, increased operating costs, reduced profits and income, affecting their long-term development. The research object of this paper is a fashion leisure clothing company with a first-line brand in China. The short life cycle is more prominent in fashion clothing. Combined with the product characteristics and sales data of the clothing company, this paper solves the difficult problems in the forecast of the sales volume of the clothing company by mathematical method, and looks for a reliable forecast scheme to improve the forecast precision of fashion clothing sales. The main research achievement of this paper is to use MATLAB as a programming tool to divide the life cycle of clothing company products and to design a non-parametric kernel density estimation model to predict the product life cycle. The non-parametric kernel estimation function is used to predict the daily and total sales of clothing in each category of the company, and the accuracy of the forecast results is tested. On the basis of the effective result, the paper puts forward some suggestions on the decision-making of the company's order according to the forecast. The research in this paper can help garment companies to establish and perfect the forecasting system, and provide reference for enterprises to arrange the production replenishment plan reasonably.
【学位授予单位】：浙江工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F274;F426.86;F224

【相似文献】