Scribd
Upload
37 views

Timeseries Forecasting Exercise-3-Sheets

The document contains seasonal indices and forecasts for visitor data from 2014 to 2021. It shows the actual quarterly visitor counts alongside seasonal indices, irregular components, forecasts and seasonal indices for each quarter. It also displays a line chart of the actual data and forecast over time, and includes an exponential smoothing model used to generate the forecasts with parameters.

Uploaded by

alyaalderazi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as XLSX, PDF, TXT or read online on Scribd
37 views

Timeseries Forecasting Exercise-3-Sheets

The document contains seasonal indices and forecasts for visitor data from 2014 to 2021. It shows the actual quarterly visitor counts alongside seasonal indices, irregular components, forecasts and seasonal indices for each quarter. It also displays a line chart of the actual data and forecast over time, and includes an exponential smoothing model used to generate the forecasts with parameters.

Uploaded by

alyaalderazi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as XLSX, PDF, TXT or read online on Scribd
You are on page 1/ 10

Seasonal Indices Forecasting DEV

demand 1 2 3
4 period Seasonal
Data centered Irregular Seasonal
YEAR Quarter Quarter_Q
n Visitors moving Componen Forecast Qn Index
averages ts
2014 Winter 1 86 1 1.1559 1
2014 Spring 2 62 68 0.9185 82 2 0.9889 2
2014 Summer 3 28 73 0.3862 52 3 0.6249 3
2014 Fall 4 94 78 1.2129 90 4 1.0915 4
2015 Winter 1 106 83 1.2848 102 5
2015 Spring 2 82 88 0.9371 87 6
2015 Summer 3 48 96 0.5000 55 7
2015 Fall 4 114 106 1.0806 96 8
2016 Winter 1 140 114 1.2281
2016 Spring 2 120 124 0.9677
2016 Summer 3 82 136 0.6029
2016 Fall 4 154 149 1.0336
2017 Winter 1 188 161 1.1713 350
2017 Spring 2 172 172 1.0029
300
2017 Summer 3 128 177 0.7252
2017 Fall 4 198 184 1.0761 250
2018 Winter 1 208 191 1.0919
200
2018 Spring 2 202 196 1.0306
2018 Summer 3 154 196 0.7857 150
2018 Fall 4 220 196 1.1224
2019 Winter 1 208 196 1.0612 100

2019 Spring 2 202 196 1.0306 50


2019 Summer 3 154 206 0.7494
2019 Fall 4 220 215 1.0233 0
1 2 3 4 1 2 3 4 1
2020 Winter 1 246 224 1.0982
2020 Spring 2 240 232 1.0345
2020 Summer 3 190 because one it will include all needed data (last-2)
2020 Fall 4 252
2021 Winter 1 293
2021 Spring 2 251
2021 Summer 3 159
2021 Fall 4 277
5
FITn = Fn+ Tn
trend
Fn = FITn-1 + α*(Dn-1 - FITn-1)
4 Tn= Tn-1 + α* β *(Dn-1 - FITn-1)
AVG FIT= SD
Year Data Fn Tn FITn FIT / no. of (Fn-Dn)^2
Qrs

2014 270 250 80 330 83 Alpha 0.8


2015 350 282 70 352 88 Beta 0.2
2016 496 350 70 420 105
2017 686 481 82 563 141 MSE
2018 784 661 102 763 191 = √ Average((F1-D1)^2.. (Fn - Dn )^2)
2019 784 780 105 885 221
2020 928 804 89 893 223
2021 ? 921 95 1016 254
Avg last n 254

Visitors

3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4

Data Forecast
Visitors
e all needed data (last-2)
ge((F1-D1)^2.. (Fn - Dn )^2)
Seasonal Indices Forecasting
1 2 3
4 period Seasonal
Quarter Data centered Irregular Seasonal
YEAR Quarter (Qn) Visitors moving Componen Forecast Qn Index
averages ts

2014 Winter 1 86 Q1
1.155927 1
2014 Spring 2 62 68 0.9185 87 Q2
0.988861 2
2014 Summer 3 28 73 0.3862 55 Q3
0.624911 3
2014 Fall 4 94 78 1.2129 96 Q4
1.09147 4
2015 Winter 1 106 83 1.2848 102 5
2015 Spring 2 82 88 0.9371 87 6
2015 Summer 3 48 96 0.5000 55 7
2015 Fall 4 114 106 1.0806 96 8
2016 Winter 1 140 114 1.2281 122
2016 Spring 2 120 124 0.9677 104
2016 Summer 3 82 136 0.6029 66
2016 Fall 4 154 149 1.0336 115
2017 Winter 1 188 161 1.1713 163 350
2017 Spring 2 172 172 1.0029 139
300
2017 Summer 3 128 177 0.7252 88
2017 Fall 4 198 184 1.0761 154 250
2018 Winter 1 208 191 1.0919 221
200
2018 Spring 2 202 196 1.0306 189
2018 Summer 3 154 196 0.7857 119 150
2018 Fall 4 220 196 1.1224 208
100
2019 Winter 1 208 196 1.0612 256
2019 Spring 2 202 196 1.0306 219 50
2019 Summer 3 154 206 0.7494 138
2019 Fall 4 220 215 1.0233 241 0
1 2 3 4 1 2 3 4 1
2020 Winter 1 246 224 1.0982 258
2020 Spring 2 240 232 1.0345 221
2020 Summer 3 190 140
2020 Fall 4 252 244
2021 Winter 1 293 293
2021 Spring 2 251 251
2021 Summer 3 159 159
2021 Fall 4 277 277
5
FITn = Fn+ Tn
Fn = FITn-1 + α*(Dn-1 - FITn-1)
4 Tn= Tn-1 + α* β *(Dn-1 - FITn-1)

SD
Year Data Fn Tn FITn AVG FIT (F1-D1)^2

Alpha
2014 270 250 80 330 83 3600 0.8
Beta
2015 350 282 70 352 88 6 0.2

2016 496 350 70 420 105 5701


MSE
2017 686 481 82 563 141 15130 71.99
2018 784 661 102 763 191 434 = √ Average((F1-D1)^2.. (Fn - Dn )^2)
2019 784 780 105 885 221 10190
2020 928 804 89 893 223 1215
2021 921 95 1016 254
Avg last n 254

Visitors

3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4

Data Forecast
Visitors
ge((F1-D1)^2.. (Fn - Dn )^2)
Seasonal Indices Forecasting
1 2 3
4 period Seasonal
Quarter Data centered Irregular Seasonal
YEAR Quarter (Qn) Visitors moving Componen Forecast Qn Index
averages ts

2014 Winter 1 86 Q1
1.155927 1
2014 Spring 2 62 68 0.9185 87 Q2
0.988861 2
2014 Summer 3 28 73 0.3862 55 Q3
0.624911 3
2014 Fall 4 94 78 1.2129 96 Q4
1.09147 4
2015 Winter 1 106 83 1.2848 102 5
2015 Spring 2 82 88 0.9371 87 6
2015 Summer 3 48 96 0.5000 55 7
2015 Fall 4 114 106 1.0806 96 8
2016 Winter 1 140 114 1.2281 122
2016 Spring 2 120 124 0.9677 104
2016 Summer 3 82 136 0.6029 66 MSE
2016 Fall 4 154 149 1.0336 115 71.99
2017 Winter 1 188 161 1.1713 163 0.1
2017 Spring 2 172 172 1.0029 139 0.2
2017 Summer 3 128 177 0.7252 88 0.3
2017 Fall 4 198 184 1.0761 154 0.4

Alpha
2018 Winter 1 208 191 1.0919 221 0.5
2018 Spring 2 202 196 1.0306 189 0.6
2018 Summer 3 154 196 0.7857 119 0.7
2018 Fall 4 220 196 1.1224 208 0.8
2019 Winter 1 208 196 1.0612 256 0.9
2019 Spring 2 202 196 1.0306 219
2019 Summer 3 154 206 0.7494 138
2019 Fall 4 220 215 1.0233 241
2020 Winter 1 246 224 1.0982 258
2020 Spring 2 240 232 1.0345 221
2020 Summer 3 190 140
2020 Fall 4 252 244
2021 Winter 1 293 293
2021 Spring 2 251 251
2021 Summer 3 159 159
2021 Fall 4 277 277
5
FITn = Fn+ Tn
Fn = FITn-1 + α*(Dn-1 - FITn-1)
4 Tn= Tn-1 + α* β *(Dn-1 - FITn-1)

SD
Year Data Fn Tn FITn AVG FIT (F1-D1)^2

Alpha
2014 270 250 80 330 83 3600 0.8
Beta
2015 350 282 70 352 88 6 0.2

2016 496 350 70 420 105 5701


MSE
2017 686 481 82 563 141 15130 71.99
2018 784 661 102 763 191 434 = √ Average((F1-D1)^2.. (Fn - Dn )^2)
2019 784 780 105 885 221 10190
2020 928 804 89 893 223 1215
2021 921 95 1016 254
Avg last n 254

Beta
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
ge((F1-D1)^2.. (Fn - Dn )^2)
1. Calculate Centered Moving Average
(N) Number of moving periods= no. of Quarters
Forecasting Starts From Period N/2
2. Find the Seasonal Irregular Component
Seasonal Irregular Components = Data / Centered Moving Average
3. Find the seasonal Indices by Averaging the cyclical components for each cycle
4. Use any forecasting technique to find the last year forecast
Avg forecast last year = last year forecast / no. of Quarters
5. Use the seasonal indices to calculate the each quarter forecast
Datan = Avg last year * Qn index

You might also like