DEMAND ANALYSIS

AND TIME SERIES FORECAST

- ARKAPRAVA GHOSH
- VIKASH PRAKASH RAJDEV
CAAS: Demand Analysis and Time Series Forecast

C R O S S AT L A N T I C AI R L I N E S ERV I C ES H AS
O VE R T H E Y EA R S H AS E XP ER I E N CE D
G RO W T H I N T E R M S O F VO L UM E O F
A I R L I NE T R AV E L . W E AT T E M P T TO
A N A LY Z E T H E S AM E A N D F OR E CA S T
T H E G R O W T H I N T ER M S O F V OL UM E F O R
O NE Y E A R .
 Dataset provided to us is a time series data indicating volume
of airline travel from 1949-60.

 Time series analysis has been performed on this dataset and

eventually we derive forecasted values of airline travel for the
corresponding year.

 Analysis procedure are based on Nelson & Plosser (1982) Time

Series Analysis

 SAS 9.2 have been used for forecasting the airline travel data.
 Data Treatment and Analysis

Augmented Dickey-Fuller Unit Root Tests (Original Data)

Type Lags Rho Pr < Rho Tau Pr < Tau F Pr > F

Zero Mean 0 0.2095 0.7312 0.91 0.9030

1 0.1910 0.7266 0.67 0.8602
2 0.1898 0.7263 0.80 0.8834
Single Mean 0 -5.2262 0.4082 -1.82 0.3714 2.22 0.5053
1 -7.3409 0.2474 -2.02 0.2784 2.39 0.4612
2 -5.2114 0.4095 -1.65 0.4545 1.80 0.6132
Trend 0 -41.8326 0.0005 -4.85 0.0006 11.86 0.0010
1 -103.902 0.0001 -7.00 <.0001 24.51 0.0010

2 -147.305 0.0001 -6.71 <.0001 22.56 0.0010

Augmented Dickey-Fuller Unit Root Tests (After Log Transformation)

Type Lags Rho Pr < Rho Tau Pr < Tau F Pr > F

Zero Mean 0 0.2095 0.7312 0.91 0.9030

1 0.1910 0.7266 0.67 0.8602
2 0.1898 0.7263 0.80 0.8834
Single Mean 0 -5.2262 0.4082 -1.82 0.3714 2.22 0.5053
1 -7.3409 0.2474 -2.02 0.2784 2.39 0.4612
2 -5.2114 0.4095 -1.65 0.4545 1.80 0.6132
Trend 0 -41.8326 0.0005 -4.85 0.0006 11.86 0.0010
1 -103.902 0.0001 -7.00 <.0001 24.51 0.0010
2 -147.305 0.0001 -6.71 <.0001 22.56 0.0010
 Data has auto-correlation and seasonality, so suitable adjusting
has been done.

 Best model fit for the time series data is ARMA(1,0). Model
selection on the basis of the MAPE criterion and MINIC option
on PROC ARIMA.

 Model has been validated on the basis of existing data only and
is found to have good predictive ability.
 Seasonality
Autocorrelations

Lag Covariance Correlation -1 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 1 Std Error

0 0.0020860 1.00000 | |********************| 0

1 -0.0007116 -.34112 | *******| . | 0.087370
2 0.00021913 0.10505 | . |** . | 0.097006
3 -0.0004217 -.20214 | ****| . | 0.097870
4 0.00004456 0.02136 | . | . | 0.101007
5 0.00011610 0.05565 | . |* . | 0.101042
6 0.00006426 0.03080 | . |* . | 0.101275
7 -0.0001159 -.05558 | . *| . | 0.101347
8 -1.5867E-6 -.00076 | . | . | 0.101579
9 0.00036791 0.17637 | . |**** | 0.101579
10 -0.0001593 -.07636 | . **| . | 0.103891
11 0.00013431 0.06438 | . |* . | 0.104318
12 -0.0008065 -.38661 | ********| . | 0.104621
13 0.00031624 0.15160 | . |*** . | 0.115011
14 -0.0001202 -.05761 | . *| . | 0.116526
15 0.00031200 0.14957 | . |*** . | 0.116744
16 -0.0002898 -.13894 | . ***| . | 0.118197
17 0.00014703 0.07048 | . |* . | 0.119438
18 0.00003261 0.01563 | . | . | 0.119755
19 -0.0000221 -.01061 | . | . | 0.119770
20 -0.0002435 -.11673 | . **| . | 0.119777
21 0.00008042 0.03855 | . |* . | 0.120643
22 -0.0001906 -.09136 | . **| . | 0.120737
23 0.00046574 0.22327 | . |****. | 0.121263
24 -0.0000384 -.01842 | . | . | 0.124362
 The ARIMA Procedure

Inverse Autocorrelations

Lag Correlation -1 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 1

1 0.32632 | . |******* |
2 0.09594 | . |**. |
3 0.09992 | . |**. |
4 0.10889 | . |**. |
5 -0.12127 | .**| . |
6 -0.16601 | ***| . |
7 -0.05979 | . *| . |
8 0.02949 | . |* . |
9 -0.08480 | .**| . |
10 0.01413 | . | . |
11 0.10508 | . |**. |
12 0.37985 | . |******** |
13 0.12446 | . |**. |
14 0.05655 | . |* . |
15 0.05144 | . |* . |
16 0.09499 | . |**. |
17 -0.04675 | . *| . |
18 -0.05409 | . *| . |
19 0.04468 | . |* . |
20 0.03780 | . |* . |
21 -0.06184 | . *| . |
22 -0.00864 | . | . |
23 -0.06173 | . *| . |
24 0.04423 | . |* . |
Partial Autocorrelations

Lag Correlation -1 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 1

1 -0.34112 | *******| . |
2 -0.01281 | . | . |
3 -0.19266 | ****| . |
4 -0.12503 | ***| . |
5 0.03309 | . |* . |
6 0.03468 | . |* . |
7 -0.06019 | . *| . |
8 -0.02022 | . | . |
9 0.22558 | . |***** |
10 0.04307 | . |* . |
11 0.04659 | . |* . |
12 -0.33869 | *******| . |
13 -0.10918 | .**| . |
14 -0.07684 | .**| . |
15 -0.02175 | . | . |
16 -0.13955 | ***| . |
17 0.02589 | |* . |
18 0.11482 | |**. |
19 -0.01316 | | . |
20 -0.16743 | ***| . |
21 0.13240 | . |*** |
22 -0.07204 | . *| . |
23 0.14285 | . |*** |
24 -0.06733 | . *| . |
Autocorrelation Check for White Noise
To Chi- Pr >
Lag Square DF Chi Sq --------------------Autocorrelations--------------------

6 23.27 6 0.0007 -0.341 0.105 -0.202 0.021 0.056 0.031

12 51.47 12 <.0001 -0.056 -0.001 0.176 -0.076 0.064 -0.387
18 62.44 18 <.0001 0.152 -0.058 0.150 -0.139 0.070 0.016
24 74.27 24 <.0001 -0.011 -0.117 0.039 -0.091 0.223 -0.018

Conditional Least Squares Estimation

Standard Approx
Parameter Estimate Error t Value Pr > |t| Lag

MU 0.0002569 0.0027959 0.09 0.9269 0

AR1,1 -0.34598 0.08852 -3.91 0.0001 1
AR1,2 -0.01376 0.08894 -0.15 0.8773 2

Constant Estimate 0.000349

Variance Estimate 0.001886
Std Error Estimate 0.043428
AIC -447.073
SBC -438.447
Number of Residuals 131
* AIC and SBC do not include log determinant.
Correlations of Parameter Estimates
Parameter MU AR1,1 AR1,2

MU 1.000 -0.007 -0.013

AR1,1 -0.007 1.000 0.345
AR1,2 -0.013 0.345 1.000

Autocorrelation Check of Residuals

To Chi- Pr >
Lag Square DF Chi Sq --------------------Autocorrelations--------------------

6 7.98 4 0.0925 -0.003 -0.067 -0.210 -0.029 0.086 0.039

12 39.75 10 <.0001 -0.055 0.046 0.195 -0.009 -0.098 -0.407
18 43.74 16 0.0002 0.020 0.039 0.115 -0.089 0.043 0.040
24 56.65 22 <.0001 -0.053 -0.138 -0.030 -0.013 0.239 0.026

Model for variable lair

Estimated Mean 0.000257

Period(s) of Differencing 1,12

Autoregressive Factors

Factor 1: 1 + 0.34598 B**(1) + 0.01376 B**(2)

 Forecasts for variable lair

Obs Forecast Std Error 95% Confidence Limits

145 6.1021 0.0434 6.0170 6.1872

146 6.0367 0.0519 5.9350 6.1384

147 6.1065 0.0615 5.9860 6.2270

148 6.2021 0.0692 6.0666 6.3377

149 6.2260 0.0762 6.0767 6.3754

150 6.3516 0.0826 6.1896 6.5135

151 6.5025 0.0886 6.3289 6.6761

152 6.4767 0.0942 6.2921 6.6612

153 6.3005 0.0994 6.1057 6.4954

154 6.2037 0.1044 5.9990 6.4084

155 6.0367 0.1092 5.8227 6.2508

156 6.1393 0.1138 5.9162 6.3623

Obs DATE lair FORECAST STD L95 U95 RESIDUAL Counter

1 JAN60 6.03309 6.05991 0.04343 5.97479 6.14503 -0.02682 133

2 FEB60 5.96871 5.99449 0.04343 5.90937 6.07960 -0.02578 134
3 MAR60 6.03787 6.14563 0.04343 6.06051 6.23075 -0.10776 135
4 APR60 6.13340 6.04888 0.04343 5.96376 6.13400 0.08452 136
5 MAY60 6.15698 6.15232 0.04343 6.06720 6.23744 0.00466 137
6 JUN60 6.28227 6.28459 0.04343 6.19948 6.36971 -0.00233 138
7 JUL60 6.43294 6.42943 0.04343 6.34432 6.51455 0.00351 139
8 AUG60 6.40688 6.45257 0.04343 6.36745 6.53769 -0.04569 140
9 SEP60 6.23048 6.23468 0.04343 6.14956 6.31980 -0.00420 141
10 OCT60 6.13340 6.09839 0.04343 6.01327 6.18351 0.03501 142
11 NOV60 5.96615 6.00540 0.04343 5.92028 6.09052 -0.03925 143
12 DEC60 6.06843 6.09563 0.04343 6.01051 6.18075 -0.02720 144
13 JAN61 . 6.10211 0.04343 6.01699 6.18723 . 145
14 FEB61 . 6.03667 0.05189 5.93496 6.13837 . 146
15 MAR61 . 6.10649 0.06150 5.98595 6.22702 . 147
16 APR61 . 6.20215 0.06915 6.06662 6.33768 . 148
17 MAY61 . 6.22602 0.07621 6.07665 6.37539 . 149
18 JUN61 . 6.35156 0.08262 6.18962 6.51349 . 150
19 JUL61 . 6.50249 0.08858 6.32887 6.67611 . 151
20 AUG61 . 6.47669 0.09416 6.29213 6.66124 . 152
21 SEP61 . 6.30055 0.09943 6.10566 6.49543 . 153
22 OCT61 . 6.20372 0.10444 5.99903 6.40841 . 154
23 NOV61 . 6.03672 0.10921 5.82267 6.25078 . 155
24 DEC61 . 6.13926 0.11379 5.91624 6.36228 . 156
 Conclusion

Results

 Time series data predicts that volume of airline travel increases

over the successive 12 months.

 Prediction for the entire duration i.e., actual and the forecasted
value has been given.