spss销量预测.doc

1、案例11 预测食品和饮料的销售量Vintage饭店位于靠近佛罗里达的Fort Myers的Captiva岛上，是一个公众常去的场所。它由Karen Payne拥有和经营，到目前经营已超过30年。在这期间，Karen一直在寻求店建立以新鲜海味设置的高质量正餐的饭店信誉。Karen及其员工的努力被证实是成功的，她的饭店成为岛上最好的营业额增长最快的饭店之一。Karen为确定饭店未来的增长计划，需要建立一个系统，这个系统可使她提前一年预测今后每个月食品和饮料的销售量。Karen拥有如下资料，这些资料是在三年的经营中有关食品和饮料的销售总量：月份123456789101112第一年2422352321

2、78184140145152110130152206第二年263238247193193149157161122130167230第三年282255265205210160166174126148173235管理报告分析Vintage饭店的销售资料，为Karen准备一份囊括你的发现、预测和建议的报告。其内容包括：（1）时间数列的图形。（2）对数据的季节性分析。指出每个月的季节指数，并讨论各月销售量的高低。季节指数有直观上的意义吗？对此应加以讨论。（3）预测第四年各月的销售量。（4）当用来说明新的销售资料时，提出对你所建立的系统的建议。（5）在你报告的附录中，给出评论分析的结果。假设第四年1月份

3、的销售额为295000美元，你的预测误差为多少？如果这个误差太大，Karen可能会对你的预测值和实际销售额的差异产生疑虑，你将如何消除她对预测方法的疑虑？解决方案一答：（1）Vintage饭店的销售的时间序列图形如下所示。（2）季节分析如下图所示。其中SALE-SA为季节调整后的销售量；SALE-TC为季节调整后的销售量趋势；SALE-SF为季节指数；SALE-IR为不规则因素。obsSALE_SASALE_TCSALE_SFSALE_IR2001M01167.7192171.93141.4428880.9755012001M02180.4607172.2091.3022231.0479162

4、001M03172.5848172.6991.3442670.9993382001M04171.7631173.30681.0363110.9910932001M05175.0472174.04941.0511451.0057332001M06175.1375174.9630.7993721.0009982001M07175.1497175.94490.8278630.995482001M08177.5969176.90520.8558711.003912001M09177.5796177.79410.619440.9987942001M10179.5815178.6110.7239051.0

5、054342001M11178.6836179.52990.8506660.9952862001M12179.748180.58181.1460490.9953832002M01182.2734181.61891.4428881.0036042002M02182.7644182.67361.3022231.0004972002M03183.7433183.70661.3442671.00022002M04186.2376184.70371.0363111.0083042002M05183.6093185.65451.0511450.9889842002M06186.3963186.66770.

6、7993720.9985462002M07189.6448187.95430.8278631.0089942002M08188.1125189.54180.8558710.9924592002M09196.9519191.25960.619441.0297622002M10179.5815192.90120.7239050.9309512002M11196.3168194.24660.8506661.0106582002M12200.6895195.23561.1460491.0279352003M01195.4414196.01791.4428880.9970592003M02195.819

7、196.63281.3022230.9958612003M03197.1335197.23671.3442670.9994762003M04197.8171198.04231.0363110.9988632003M05199.7821199.10191.0511451.0034172003M06200.1571200.31390.7993720.9992172003M07200.5162201.4290.8278630.9954682003M08203.3017202.37670.8558711.0045712003M09203.4094203.18010.619441.0011292003M

8、10204.4466203.85350.7239051.0029092003M11203.3701204.38650.8506660.9950272003M12205.0523204.8061.1460491.001203（3）预测的第四年的销量如下表所示。Dependent Variable: SALE_TCMethod: Least SquaresDate: 11/24/11 Time: 18:09Sample (adjusted): 2001M01 2003M12Included observations: 36 after adjustmentsVariableCoefficientS

9、td. Errort-StatisticProb.T1.0322110.01335577.288790.0000C169.01710.283360596.47450.0000R-squared0.994340Mean dependent var188.1130Adjusted R-squared0.994174S.D. dependent var10.90592S.E. of regression0.832430Akaike info criterion2.525017Sum squared resid23.55993Schwarz criterion2.612990Log likelihoo

10、d-43.45030Hannan-Quinn criter.2.555722F-statistic5973.557Durbin-Watson stat0.164544Prob(F-statistic)0.000000SALE_TC = 1.03221113243*T + 169.017061159 其中T=1时为第一年一月。DATESALESALE_TCSALE_TCFSALE_SFSALE_F2001M01242171.9314170.04931.442888245.36212001M02235172.209171.08151.302223222.78622001M03232172.6991

11、72.11371.344267231.36682001M04178173.3068173.14591.036311179.4332001M05184174.0494174.17811.051145183.08652001M06140174.963175.21030.799372140.05822001M07145175.9449176.24250.827863145.90472001M08152176.9052177.27480.855871151.72432001M09110177.7941178.3070.61944110.45052001M10130178.611179.33920.72

12、3905129.82452001M11152179.5299180.37140.850666153.43582001M12206180.5818181.40361.146049207.89742002M01263181.6189182.43581.442888263.23442002M02238182.6736183.4681.302223238.91632002M03247183.7066184.50021.344267248.01762002M04193184.7037185.53241.036311192.26932002M05193185.6545186.56471.051145196

13、.10652002M06149186.6677187.59690.799372149.95972002M07157187.9543188.62910.827863156.1592002M08161189.5418189.66130.855871162.32562002M09122191.2596190.69350.61944118.12322002M10130192.9012191.72570.723905138.79122002M11167194.2466192.75790.850666163.97262002M12230195.2356193.79011.146049222.0932003M01282196.0179194.82231.442888281.106820