案例 黑色塑料

SUMMARY OUTPUT回归统计Multiple R 0.955259R Square 0.912519Adjusted R Square0.90627标准误差 345.9009观测值16方差分析dfSSMSF Significance F回归分析1 17472676 17472676 146.0347 8.55E-09残差141675064 119647.4总计15 19147740Coefficients 标准误差t StatP-value Lower 95%Upper 95%下限 95.0%上限 95.0%Intercept 2595.071 215.1172 12.06352 8.74E-09 2133.691 3056.452 2133.691 3056.452X Variable 1226.6941 18.75911 12.08448 8.55E-09 186.4598 266.9284 186.4598 266.9284上限 95.0%SUMMARY OUTPUT回归统计Multiple R 0.539521R Square 0.291083Adjusted R Square0.251699标准误差 2717.841观测值20方差分析dfSSMSF Significance F回归分析1 54593578 54593578 7.390836 0.014083残差18 1.33E+087386658总计19 1.88E+08Coefficients 标准误差t StatP-value Lower 95%Upper 95%下限 95.0%上限 95.0%Intercept 2044.105 1262.522 1.619064 0.122821-608.356 4696.567-608.356 4696.567X Variable 1286.5233 105.3933 2.718609 0.014083 65.10012 507.9465 65.10012 507.9465上限 95.0%黑色塑料的需求量与误差分析时期实际需求量 剔除季节性影响后的需求Dt=227t+2595St=Dt/Dt季节性需求季节性需求静态方法SHEET 4122509623.448.35217379431.840.90324123576.2517901.350.77472693786.2526372.761.60535243967.87534841.01621434072.12543310.4973459412051780.67870564272.37560251.17941204237.37568720.601027664274.12577190.361125564595.12585660.301282534968.594130.881354915390.375102600.541443826083111070.391543156575.375119540.3616120356509.25128010.941756486489.5136480.411836966688.25144950.25194843153420.322013097161890.8121 Dt=227t+259522232425262728293031324期移动平均法时期实际需求量 移动平均值12250 需求水平2173732412472693417535243735.5621433837734594098.75870564045.5941204194.51027664350.251125564124.51282534423.751354914766.51443825170.51543155610.2516120356555.7517564865951836966423.51948436555.520130976821212223242526272829303132时期实际需求量 需求水平预测需求单一指数平滑法阿尔法=0.105052.6122504772.345052.6217374468.8064772.34324124263.12544468.806472694563.712864263.1254535244459.7415744563.71286621434228.0674174459.741574734594151.1606754228.067417870564441.6446074151.160675941204409.4801474441.6446071027664245.1321324409.4801471125564076.2189194245.1321321282534493.8970274076.2189191354914593.6073244493.8970271443824572.4465924593.6073241543154546.7019334572.44659216120355295.5317394546.7019331756485330.7785655295.5317391836965167.3007095330.7785651948435134.8706385167.30070920130975931.0835745134.870638215931225931235931245931255931265931275931285931295931305931315931325931S1S2S3S40.470.681.171.66时期实际需求量 需求水平L需求趋势T预测需求需求趋势和季节性需求修正后的指数平滑法（winter）阿尔法=0.1，贝塔=0.1，伽马=0.102595227122503018.523404246.6523404 23562.3038.35217373308.232638250.9580298 2933.36330.90324123716.463091266.685272 2745.19610.77472695131.429271381.5133629 6378.91651.60535245711.435605401.3626599 46030.343621435957.475885385.830422 5491.6057734596444.933123395.9931036 4892.5785870567658.1102477.7115009 10955.579941208198.835275484.0128584 67930.0851027668403.073959456.0354409 7800.4829 1125568517.028247421.82732566833.0121282539800.927462508.0345145 14315.36313549110446.36365521.774682 86074.73114438210803.66493505.3273412 9853.538215431511096.17815484.0459292 8931.8019161203512982.83997624.307518221791.5217564813448.13486608.4062562 113612.9518369613437.26999546.4791428 12628.09219484313615.79975509.6842047 10894.955201309715499.5313647.0889397 25858.42521 Dt=2595+227t86074.731229717.7557238734.97212418950.37125103042.082611543.3852710302.3562822204.79129120009.433013369.0133111869.743225459.211时期实际需求量 需求水平需求趋势预测需求趋势修正后的指数平滑法（holt）阿尔法=0.1，贝塔=0.1sheet 502044287122502322.9286.192331217372521.881277.46912609.09324122760.61509273.595599 2799.3501472693457.68962315.9434921 3034.2107535243748.669801313.447161 3773.6331621433870.205266294.25599144062.117734594093.915131287.2013788 4164.4613870564648.604859313.9502137 4381.1165941204878.299566305.524663 4962.55511027664942.041806281.3464207 5183.82421125564956.649404254.6725384 5223.38821282535515.489748285.089319 5211.32191354915769.62116281.9935283 5800.57911443825884.65322265.2973814 6051.61471543155966.455541246.9478754 6149.950616120356795.563075305.1638413 6213.40341756486955.454224290.6365721 7100.72691836966891.081717255.1356641 7246.09081948436915.895643232.1034903 7146.217420130977742.89922291.593499 7147.999121 Dt=2044+287t5800.5791226085.6684236370.7577246655.847258321.3823268626.5461278931.71289236.87382910367.2413010658.8343110950.4283211242.021黑色塑料预测需求的误差预测方法MADMAPE（%）TS的变动范围四期移动平均法2070.078125 36.398292-4.712.69单一指数平滑法225051.000013-3.914.98Holt模型214944.076572高估Winter模型19512431.45099-2.273静态方法的路径系数在+6以外，会高估需求高估结论：采用四期移动平均法2001年第一季度68212001年第二季度68212001年第三季度68212001年第四季度68212002年第一季度68212002年第二季度68212002年第三季度68212002年第四季度68212003年第一季度68212003年第二季度68212003年第三季度68212003年第四季度6821误差误差的的标准差2588预测值E=预测值-需求值At=ABS(E)MSEMADE/D802-1448144820980002098000144864.37544847-8908907917961444898116951.227931381-1031103110637381317845112342.760234223-3046304692775543307772160441.902712909025566255666536037831333669746396725.47323891174817483055073111648324562281.56214399453553528599095739419489515.46055964925932593672309484612379460736.747355737853258532582836393183390365801100131292.665693541694169173770813530669299428150.708660740514051164100943224617628939158.4874150756822682246535505299467908876382.657148566680175801756428018224770894855142561460.1159978559655963131844571806796913638127.71089220490549052405995267180076813055113.67542050084658465716645966342922571276970.34063113954108306108306117301996121286992689183881917.6130229326932686974172122032499417885252.3267118336990699048863618115866913217312144.337225926128291282916459081411089652161708797.9559861469误差的标准差21359.326818602812881690507634672914335766328449742915789预测值FE=预测值-需求值At=ABS(E)MSEMADE/D%3417-1071071144911449107 3.0363223735.51592.51592.52536056.251273752.625849.75 74.311713837378378142884896796.4167692.5 10.928014098.75-2957.252957.258745327.5632858929.203 1258.688 41.911144045.5-74.574.55550.252288253.4131021.85 1.8082524194.51428.51428.52040612.252246979.885 1089.625 51.644974350.251794.251794.253219333.0632385887.482 1190.286 70.197574124.5-4128.54128.517044512.254218215.578 1557.563 50.024234423.75-1067.251067.251139022.5633876083.021 1503.083 19.436354766.5384.5384.5147840.253503258.744 1391.225 8.7745325170.5855.5855.5731880.253251315.244 1342.523 19.826195610.25-6424.756424.7541277412.566420156.688 1766.042 53.383886555.75907.75907.75824010.06255989683.87 1700.019 16.0720665952899289984042016162149.379 1785.661 78.436156423.51580.51580.52497980.255917871.438 1771.983 32.634736555.5-6541.56541.542791222.258222455.863 2070.078 49.946556821误差的标准差2587.59868216821682168216821682168216821682168216821E=预测需求-需求值At=ABS(E)MSEMADE/D28032803785456778545672803124.5630353035921328985339282919174.74612057205742304517099436263285.27388-3006300690352827583397272541.351971040104010810036282918238829.5037723172317536729261303142376108.10747697695914655339050214722.23381-2905290584380925726430224141.16836322322103455510165520287.8069081643164327010274861592199059.417211689168928531674679008196266.08498-41774177174455005742882214750.60925-9979979942145377600205818.1588621221244778499668419264.82901225725766279466799118155.966317-74887488560746087880904217062.22101-352352124234742463020636.2405851635163526725017160622203944.23102324324105171678928319496.696277-79627962633955049619594225060.79354误差的标准差2811.884E=预测需求-需求值At=ABS(E)MSEMADE/D213122131245421428045421428021312947.21351196119614312852278227821125468.87526333333111020151918862761413.8141-890890792249114137208593312.2449242506425061806789180452667603132481206.196334933491121316037909186211598156.25791434143420551473252294741014641.444883900390015206718286476630936555.2661563810638104071726954707059999154141548.788503450342534601863888860114376182.0131427742771829283258247080413458167.332260626062367522425369942571284273.4564780584805846493737692995205291180531467.56547254722993773092625760817154124.8639461746172131486086592809116318106.994397579757951896888177569411590881.06789107965107965116564299451455325941213231911.5618932893279782276137890684920635241.66926052605236626164130826049719867124.9629127611276116285396712509901701951297.43777误差的标准差24390.19E=预测需求-需求值At=ABS(E)MSEMADE/D818165616561813.687287276054138355147750.2066838738715004030571444716.05929-42354235179334414712646139458.2582125025062317378258011657.08381919191936830103765985129189.552827057054976763299084120720.39495-2675267571550023781073139137.909358438437098993439832133020.450372418241858458743680436143887.412326672667711496039926651550104.3579-3042304292518054430927167436.8554231031095839409745915695.6379361670167027876134003898157738.101661835183533670443961441159442.52493-58225822338909875832038185848.372221453145321104155613119183425.7210935503550126031456001454193096.052242303230353048105964788194947.55766-59495949353906117436079214945.42262误差的标准差2686.534MAPEbiasTS=bias/MAD64.37544-1448-157.80169-2338-252.78787-3370-350.06658-6416-4185.1479191502.99396087167.8836208983.71749901146.1089214334.37864418132.4387240265.21492325261.3527772837.7185193250.2883814528.63908662241.9427855039.56466536228.66899232510.5357586323.395617250012.0999625309.418117809613.0591393296.368618300114.0171844282.241819146614.9951034378.439429977316.3022184371.433130909917.2825755359.480731608918.2588124346.404432891819.2490913高估MAPE%biasTS=bias/MAD3.036322-107-138.674021485.5 1.7481612229.425351863.5 2.6909747332.5468-1093.75-0.868960726.39909-1168.25-1.143269630.60674260.25 0.2388436436.262572054.5 1.7260561737.98278-2074-1.331567835.92206-3141.25-2.089870833.20731-2756.75-1.981527131.99084-1901.25-1.416177133.7736-8326-4.714498132.41194-7418.25-4.363627135.69938-4519.25-2.530855935.49507-2938.75-1.658452436.39829-9480.25-4.57965812.69097473-4.7144981MAPEbiasTS=bias/MAD124.5628031149.653158382128.19337895310648891.793979119159292.4825873993.9238682453.4699933183.6824290144.1993980178.3681661102.7258089970.5280264313.171093669.4169480754.0584215169.1140497644.9757114467.5719755872.6024495563.7709645902.2298581559.5608248022.492386275650592.7869919956-2429-1.119614853.4282-2782-1.348496852.91725-1147-0.562507750.48456-823-0.422151151.00001-8785-3.9052261-3.90522614.97571144MAPEbiasTS=bias/MAD947.2135213121508.0444225092343.3009228423261219523.69995433450644584.86562403400.7669678075.84650843349.4352692406.82454118312.6641731407.8099129450.01121369508.88451318423.21141419849.8761749399.949714626110.8677275372.741915232411.8613793456.958723290812.9013916433.237623837913.896225841124299614.890910539125275215.8880981480.291236071716.9165157467.034436964917.9137306449.030637570118.9103978431.45138846319.9087662高估MAPEbiasTS=bias/MAD3.681126.90334953223.288661340332-2894-2.076577727-2645-2.270196237.46013-726-0.562190935.02225-20-0.016683435.38314-2695-1.938110633.72394-1852-1.39319539.092785650.3930271745.0259732332.0853868444.345091910.1141058841.367625010.318987541.13434-1062-0.67389144140052.5128995542-1816-0.977552140.73523-364-0.198259443.808431861.6513665244.0057354902.8163043444.07657-459-0.2137279-2.27019623