Quantitative and Qualitative Forecasting

Template for Example 18.4

Forecasting		Simple linear regression
Data Elissa Torres: Forecasting: Submodel = 15; Problem size @ 8 by 2				Forecasts and Error Analysis						Tracking Signal
Period	Demand (y)	Period(x)		Forecast	Error	Absolute	Squared	Abs Pct Err		Cum error	Cum Abs Err	Mad	Track Signal (Cum error/MAD)
2011 Q1	300	228.3		228.31	71.69	71.69	5139.33	0.24
2011 Q2	200	280.6		280.61	-80.61	80.61	6497.62	0.40		-80.61	80.61	80.61	-1.00
2011 Q3	220	332.9		332.90	-112.90	112.90	12747.46	0.51		-193.51	193.51	88.40	-2.19
2011 Q4	530	385.1		385.10	144.90	144.90	20995.56	0.27		-48.61	338.41	102.52	-0.47
2012 Q1	520	437.4		437.40	82.60	82.60	6823.02	0.16		33.99	421.01	98.54	0.34
2012 Q2	420	489.6		489.60	-69.60	69.60	4843.51	0.17		-35.61	490.61	93.72	-0.38
2012 Q3	400	541.9		541.89	-141.89	141.89	20133.40	0.35		-177.50	632.50	100.60	-1.76
2012 Q4	700	594.2		594.19	105.81	105.81	11195.95	0.15		-71.69	738.31	101.25	-0.71
				Total	-0.00	810.00	88375.84	2.26
Intercept	0.02			Average	-0.00	101.25	11046.98	0.28
Slope	1.00				Bias	MAD	MSE	MAPE
						SE	121.36
Forecast	9.02	9
Using Linear Regression Method						Correlation	0.75
						Coefficient of determination	0.56
Quarter	Actual Amount	Trend from forecast	Ration of Actual/Trend			Seasonal Factor(Av. Of Same Qtr for 2011 and 2012)
2011
1	300	228.31	1.31		1	1.25
2	200	280.61	0.71		2	0.79
3	220	332.90	0.66		3	0.70
4	530	385.10	1.38		4	1.28
2012
1	520	437.40	1.19
2	420	489.60	0.86
3	400	541.89	0.74
4	700	594.19	1.18
Forecast Including Trends		Intercept=176.1, Slope=52.3
FITSt	=	FIT X Seasonal
I-2013 FITS9	9	11.29
I-2013 FITS10	10	7.87
I-2013 FITS11	11	7.71
I-2013 FITS12	12	15.36

Regression

228.3 280.60000000000002 332.9 385.1 437.4 489.6 541.9 594.20000000000005 300 200 220 530 520 420 400 700

If this is trend analysis then simply enter the past demands in the demand column. If this is causal regression then enter the y,x pairs with y first and enter a new value of x at the bottom in order to forecast y.