發(fā)布時(shí)間：2018-06-27 22:45

  本文選題：決策支持 + 銷售預(yù)測(cè)　； 參考：《浙江理工大學(xué)》2017年碩士論文

【摘要】：隨著我國(guó)信息技術(shù)的高速發(fā)展,企業(yè)要想快速提高其在市場(chǎng)中的核心競(jìng)爭(zhēng)力,需要有效利用自身積累的海量數(shù)據(jù)來不斷科學(xué)合理的制定各種決策。因此,研究能夠輔助企業(yè)管理者進(jìn)行科學(xué)決策的決策支持系統(tǒng)顯得尤為重要。本文針對(duì)決策支持系統(tǒng)的國(guó)內(nèi)外現(xiàn)狀,對(duì)銷售預(yù)測(cè)與客戶細(xì)分兩個(gè)核心內(nèi)容進(jìn)行了研究,研究?jī)?nèi)容如下:1)銷售預(yù)測(cè)。預(yù)測(cè)是指在自身所掌握信息的基礎(chǔ)上,按照一定的方法和規(guī)律對(duì)未來進(jìn)行測(cè)算。準(zhǔn)確的銷售預(yù)測(cè)不僅有利于企業(yè)進(jìn)行科學(xué)決策,而且還有利于企業(yè)獲取最大利潤(rùn)。本文首先提出了選擇對(duì)產(chǎn)品銷量影響較大的因素作為BP神經(jīng)網(wǎng)絡(luò)模型的輸入?yún)?shù);然后針對(duì)BP模型難以確定隱含層結(jié)點(diǎn)個(gè)數(shù)、易陷入局部極值等問題,構(gòu)建了遺傳算法優(yōu)化BP網(wǎng)絡(luò)預(yù)測(cè)模型;最后,通過把本文構(gòu)建的預(yù)測(cè)模型與未優(yōu)化的BP模型進(jìn)行比較和預(yù)測(cè)分析,證明了BP神經(jīng)網(wǎng)絡(luò)在遺傳算法的優(yōu)化下能夠得到較高的預(yù)測(cè)精確率。2)客戶細(xì)分�？蛻艏�(xì)分能夠使企業(yè)更精準(zhǔn)的把握客戶群,為企業(yè)展開精準(zhǔn)化營(yíng)銷提供決策支持,穩(wěn)定核心客戶,吸引新客戶,最大限度的提高企業(yè)效益。本文從客戶忠誠(chéng)度和客戶價(jià)值兩個(gè)維度進(jìn)行客戶細(xì)分,構(gòu)建了多維交叉客戶細(xì)分模型。首先,客戶忠誠(chéng)度選擇決策樹方法基于“最近購(gòu)買時(shí)間”、“購(gòu)買金額”和“購(gòu)買頻次”,將客戶分為流失客戶、浮動(dòng)客戶與忠誠(chéng)客戶;然后,針對(duì)客戶價(jià)值本文提出了近鄰傳播(Affinity Propagation)算法優(yōu)化k-means算法對(duì)客戶進(jìn)行聚類分析,將客戶分為低價(jià)值客戶、普通客戶、主要客戶與高價(jià)值客戶;最后,對(duì)兩個(gè)維度進(jìn)行交叉獲得更精確化的客戶細(xì)分模型,并針對(duì)每一類細(xì)分客戶提供適當(dāng)營(yíng)銷策略供決策者參考。
[Abstract]:With the rapid development of information technology in our country, if enterprises want to improve their core competitiveness in the market quickly, they need to effectively use the massive data accumulated by themselves to make all kinds of decisions scientifically and rationally. Therefore, it is very important to study the decision support system which can assist enterprise managers to make scientific decision. Aiming at the present situation of decision support system at home and abroad, this paper studies the two core contents of sales forecast and customer segmentation. The research contents are as follows: 1) sales forecast. Prediction is to measure the future according to certain methods and laws on the basis of the information we have. Accurate sales forecast is not only helpful for enterprises to make scientific decision, but also for enterprises to obtain maximum profit. In this paper, the factors that have a great influence on the sales volume of the product are proposed as the input parameters of BP neural network model, and then the BP model is difficult to determine the number of hidden layer nodes and is prone to fall into local extremum. The genetic algorithm optimized BP network prediction model is constructed. Finally, the prediction model constructed in this paper is compared with the unoptimized BP model. It is proved that BP neural network can obtain a high prediction accuracy rate of .2) customer segmentation under the optimization of genetic algorithm. Customer segmentation can make the enterprise grasp the customer group more accurately, provide decision support for the enterprise to launch precision marketing, stabilize the core customers, attract new customers, and maximize the efficiency of the enterprise. Based on customer loyalty and customer value, this paper constructs a multi-dimensional cross-customer segmentation model. First, the customer loyalty selection decision tree method is based on "recent purchase time", "purchase amount" and "purchase frequency", which divides customers into lost customers, floating customers and loyal customers. In this paper, Affinity propagation algorithm is proposed to optimize the k-means algorithm for customer value clustering analysis. Customers are divided into low value customers, ordinary customers, main customers and high value customers. The two dimensions are crossed to obtain a more accurate customer segmentation model, and appropriate marketing strategies for each type of subdivision customers are provided for the reference of decision makers.
【學(xué)位授予單位】：浙江理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：F274;TP18

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 梁羅希;吳江;;決策支持系統(tǒng)發(fā)展綜述及展望[J];計(jì)算機(jī)科學(xué);2016年10期

2 王亞陽;王漢斌;;基于熵值-理想點(diǎn)法的煤礦事故應(yīng)急資源調(diào)配研究[J];安全與環(huán)境工程;2014年06期

3 金建國(guó);;聚類方法綜述[J];計(jì)算機(jī)科學(xué);2014年S2期

4 李家會(huì);;灰色預(yù)測(cè)模型在產(chǎn)品銷售預(yù)測(cè)中的應(yīng)用[J];科技信息;2014年11期

5 李帆;曲世琳;于丹;曹勇;毛曉峰;;基于運(yùn)行數(shù)據(jù)人工神經(jīng)網(wǎng)絡(luò)的空調(diào)系統(tǒng)逐時(shí)負(fù)荷預(yù)測(cè)[J];建筑科學(xué);2014年02期

6 葉明;谷晨霞;;“大數(shù)據(jù)”時(shí)代決策支持系統(tǒng)新發(fā)展[J];信息安全與技術(shù);2013年08期

7 馬永杰;云文霞;;遺傳算法研究進(jìn)展[J];計(jì)算機(jī)應(yīng)用研究;2012年04期

8 王文賢;金陽;陳道斌;;基于RFM模型的個(gè)人客戶忠誠(chéng)度研究[J];金融論壇;2012年03期

9 毛健;趙紅東;姚婧婧;;人工神經(jīng)網(wǎng)絡(luò)的發(fā)展及應(yīng)用[J];電子設(shè)計(jì)工程;2011年24期

10 劉曉勇;付輝;;一種快速AP聚類算法[J];山東大學(xué)學(xué)報(bào)(工學(xué)版);2011年04期

相關(guān)重要報(bào)紙文章 前1條

1 姜益兵;;淺析服務(wù)業(yè)之魂——客戶忠誠(chéng)度[N];中華工商時(shí)報(bào);2012年

相關(guān)碩士學(xué)位論文 前8條

1 蔣雪梅;基于改進(jìn)人工蜂群算法的聚類研究[D];大連海事大學(xué);2014年

2 劉文龍;聚類分析在港口客戶細(xì)分中的應(yīng)用[D];北京交通大學(xué);2015年

3 張雷;決策支持系統(tǒng)在企業(yè)管理中的研究與應(yīng)用[D];東北師范大學(xué);2013年

4 王云霞;基于主成分—BP神經(jīng)網(wǎng)絡(luò)的股票預(yù)測(cè)[D];山東大學(xué);2011年

5 紀(jì)云云;基于蟻群神經(jīng)網(wǎng)絡(luò)的鋁帶坯晶粒度軟測(cè)量建模研究[D];中南大學(xué);2011年

6 馬勛亮;微分進(jìn)化神經(jīng)網(wǎng)絡(luò)算法研究[D];哈爾濱工業(yè)大學(xué);2010年

7 段素花;電信運(yùn)營(yíng)中的客戶價(jià)值模型的分析與設(shè)計(jì)[D];成都理工大學(xué);2010年

8 繆志剛;基于BP神經(jīng)網(wǎng)絡(luò)的銷售預(yù)測(cè)模型[D];蘇州大學(xué);2007年

，

本文編號(hào)：2075541