马尔可夫预测法在国际服装流行色预测中的应用  

Applying Markov model to predict trend of fashion colors

作　　者：常丽霞[1,2] 高卫东[1] 张万琴[2] 马宝林[2] 吴增宝[3] 

机构地区：[1]江南大学纺织服装学院,江苏无锡214122 [2]河南科技学院,河南新乡453003 [3]四川大学,四川成都610064

出　　处：《毛纺科技》2012年第7期44-47,共4页Wool Textile Journal

摘　　要：马尔可夫(MarKov)预测法是应用概率论中的马尔可夫链的理论和方法来研究随机事件的变化,并借此分析预测未来变化趋势的一种方法。文章以国际流行色权威机构国际色彩委员会发布的2007—2011年的国际女装春夏流行色定案为分析对象,应用马尔可夫预测方法对流行色定案的色彩进行预测分析,以原始数据与预测数据的相对误差百分比探讨此方法的预测精度。结果显示:应用马尔可夫预测法对5年来色彩的预测精度稳定性不统一,但是对蓝色调的预测精度很高,达90%。Markov prediction method is the application of probability theory in Markov chain theory and method to study the changes of random events and to predict future trends of a method. This paper proposed a quantitative forecasting method to discuss the prediction of fashion colors by Markov prediction theory. It took successive 5 years data of women＇s fashion color palettes, released by international authoritative agency, as subjects and processed the forecasting through Markov state transition probability matrix of the estimation model. Results show that this method has higher prediction accuracy with accuracy 90% toward the tone of blue.

关 键 词：流行色 马尔可夫模型 转移概率矩阵 非线性归划 

分 类 号：TS94[轻工技术与工程—服装设计与工程]