Example: dental hygienist

Naïve Forecast Naï ve, & Moving Average Simple …

Na ve Forecast Year Actual Na ve Forecast error A F E=A-F. 0 10. Simplest possible Forecast Na ve, & Moving Average 1 ? ? Tomorrow will be like today 2. Simple Slope Na ve is the basis for comparison of all 3. 4. methods. 5. Ted Mitchell 6. Ignores any historical data previous to today 7. 8. Use today's result to Forecast tomorrow Year Actual Na ve Forecast error Year Actual Na ve Forecast error Year Actual Na ve Forecast error A F E=A-F A F E=A-F A F E=A-F. 0 10 0 10 0 10. 1 ? 10 1 12 10 2 1 12 10 2. 2 2 2 14 12 2. 3 3 3 15 14 1. 4 4 4 16 15 1. 5 5 5 17 16 1. 6 6 6 19 17 2. 7 7 7 21 19 2. 8 8 8 23 21 2. N=8 N=8 ? 23 SE =13. Use today's result to Forecast tomorrow Use today's result to Forecast tomorrow Use today's result to Forecast tomorrow ME = 1.

5 Do a Forecast for Period 2 0 1 2 time Sales Revenue Last period Naï ve Forecast Two Periods Ago Last period plus X Percent • The sales in the last period plus the

Tags:

  Simple, Moving, Forecast, Average, Forecast na, 239 ve, Amp moving average simple

Information

Domain:

Source:

Link to this page:

Please notify us if you found a problem with this document:

Other abuse

Transcription of Naïve Forecast Naï ve, & Moving Average Simple …

1 Na ve Forecast Year Actual Na ve Forecast error A F E=A-F. 0 10. Simplest possible Forecast Na ve, & Moving Average 1 ? ? Tomorrow will be like today 2. Simple Slope Na ve is the basis for comparison of all 3. 4. methods. 5. Ted Mitchell 6. Ignores any historical data previous to today 7. 8. Use today's result to Forecast tomorrow Year Actual Na ve Forecast error Year Actual Na ve Forecast error Year Actual Na ve Forecast error A F E=A-F A F E=A-F A F E=A-F. 0 10 0 10 0 10. 1 ? 10 1 12 10 2 1 12 10 2. 2 2 2 14 12 2. 3 3 3 15 14 1. 4 4 4 16 15 1. 5 5 5 17 16 1. 6 6 6 19 17 2. 7 7 7 21 19 2. 8 8 8 23 21 2. N=8 N=8 ? 23 SE =13. Use today's result to Forecast tomorrow Use today's result to Forecast tomorrow Use today's result to Forecast tomorrow ME = 1.

2 Period actual Forecast using Average of the total history Error Momentum Total Historical Average 1 10 Get the Forecast for the next period 2 ? 10/1 = 10 ? If things have momentum they are easier to Fi+1 = (1/n)( Ai) 3. predict. where 4. Averages are a measure of momentum Fi+1 = Forecast for next period 5. Various averages are used for prediction n = number of historical periods 6. Total historical Average Ai = sum of the actual results for each of 7. Moving averages the n historical periods 8. Weighted averages 9. period actual Forecast Error period actual Forecast Error 1 10 Get the Forecast for the next period 1 10 Get the Forecast for the next period Using Total Historical Average 2 12 10/1 = 10 2 2 12 10/1 = 10 2.

3 3 ? (10+12)/2 = 11 ? 3 14 (10+12)/2 = 11 3 Disadvantage 4 4 15 (10+12+14)/3 =12 Really lags behind a trend! because 5 5 16 (10+12+14+15)/4 = Uses all historical data 6 6 17 (10+12+14+15+16)/5 = Puts equal weight on every piece of historical 7 7 19 (10+12+14+15+16+17)/6 =14 5 information 8 8 21 (10+12+14+15+16+17+19)/7 = 9 9 23 (10+12+14+15+16+17+19+21)/8 = 2. period actual Forecast using a Moving Average on last 3 periods Error period actual Forecast using a Moving Average on last 3 periods Error Moving Average 1 10 1 10. 2 12 2 12. Pick the last n periods that are most relevant 3 14 3 14. Fi+1 = (1/n) ( Ai) 4 ? (10+12+14)/3 =12 ? 4 15 (10+12+14)/3 =12 3. where 5 5 ? (12+14+15)/3 = ? Fi+1 = the Forecast for next period 6 6.

4 N = the number of periods in the Moving 7 7. Average 8 8. ( Ai) = the sum of the last n periods 9 9. period actual Forecast using a Moving Average on last 3 periods Error period actual Forecast using a Moving Average on last 3 periods Error 1 10 1 10 Moving Average 2 12 2 12. 3 14 3 14 Still lags behind a trend 4 15 (10+12+14)/3 =12 3 4 15 (10+12+14)/3 =12 3 Puts equal weight on each of the historical 5 16 (12+14+15)/3 = 5 16 (12+14+15)/3 = results being used 6 ? (14+15+16)/3 =15 6 17 (14+15+16)/3 =15 2 Gives bias when seasonal data is involved 7 7 19 (15+16+17)/3 =16 3. 8 8 21 (16+17+19)/3 = If you want more weight on the most recent 9 9 23 (17+19+21)/3 =19 4 data you need a weighted Average 3. period actual Forecast using a WEIGHTED Moving Average on Error Weighted Moving Average last 3 periods Weighted Moving Average 1 10.

5 Three period Average with equal weight 2 12. Weighted Moving Average is better at Fjun = (Amar +Aapr + Amay ) / 3 3 14 responding to a trend because it puts more or 4 15 (2(10)+3(12)+4(14))/9 = weight on recent data and less weight on old Fjun = (3A mar +3A apr + 3 Amay ) / 9 5 16 (2(12)+3(14)+4(15))/9 = 14 2 data Weighted Average with more on May 6 17 (2(14)+3(15)+4(16))/9 = They get the appropriate weights by doing a Fjun = (2A mar +3A apr + 4 Amay ) / 9 7 19 (2(15)+3(16)+4(17))/9 = statistical fit to the data Na ve Again 8 21 (2(16)+3(17)+4(19))/9 = Fjun = (0A mar +0A apr + 9 Amay ) / 9 9 23 (2(17)+3(19)+4(21))/9 = Trends Simple Trend Projection Trends in the data are not handled well by Before the era of Simple statistical tools on Simple Growth and Slope For Moving averages or exponential smoothing every PC managers used Simple calculations Trends methods.

6 Of trends based on the na ve Forecast . The na ve Forecast is Ted Mitchell sales in next period t = sales in the last period (t-1) or Rt = Rt-1. 4. Do a Forecast for Period 2 Last period plus X Percent Do a Forecast for Period 2. Simple Percentage Sales The sales in the last period plus the Sales Projection uses the Revenue percentage growth over the last two periods Revenue same growth as between the last Two Two Last period Last period two periods Periods Periods Ago Ago Na ve Forecast Na ve Forecast g 0 1 2 time 0 1 2 time Getting the slope Example Calculate historical g g = Growth rate between 0 and 1. The percentage growth over the last two R1 = g R0. Sales Simple last period where periods = g Revenue plus percent R0 = $150 growth Projection Prediction for the last period would be Two R1 = $175 Last period uses the same Periods R1 = g R 0 then calculate g Ago slope as the last 175 = g(150) two periods Na ve Forecast We know R 1 and R0 so we can calculate g g = 175 / 150.

7 G = or growth is 117% g = R1/R0. 0 1 2 time 5. Prediction of Revenue in period 2 Prediction for R2 Prediction for R2 in period 2. (Revenue in period 2) = g (Revenue in period 1). Sales Sales R2 = gR1 Revenue Revenue Where R0 = 150 R2 = (175) = 204 R0 = 150 R2 = R1 = 175 R1 = 175. R1 = revenue in 1 = 175. g = Na ve Forecast Na ve Forecast Then = 175 = 175. R2 = (175) = g = g = 0 1 2 time 0 1 2 time The Problem with The last period result + percent Examples: Na ve Method Home Market in this example is improvement method & Last Period Plus Rate of experiencing a long run decline in sales as it Very dependent on the base used in the nears the end of the Product Life Cycle percentage. If you use the same percentage Change Method as time passes then the method inflates the Ted Mitchell forecasted values New Shoes Home Market Spring 478.

8 But it is Simple and very popular! 6. Period Actual Units Na ve Forecast error Period Actual Units Na ve Forecast error Period Actual Units Na ve Forecast error Sold Sold Sold A F E=A-F A F E=A-F A F E=A-F. 3 1,193,000 3 1,193,000 3 1,193,000. 4 1,193,000 4 1,023,000 1,193,000 170,000 4 1,023,000 1,193,000 170,000. 5 5 5 ? 1,023,000. 6 6 6. 7 7 7. Use today's result to Forecast tomorrow Use today's result to Forecast tomorrow Use today's result to Forecast tomorrow What to do Next? Last period + change % Forecasting period 5. You have two pieces of information Consider the last period plus the decline rate Sales in 5 = decline rate (sales in 4). Industry Sales in period 3 = 1,193,000 from the two previous periods Sales in 5 = (1,023,000).

9 Industry Sales in period 4 = 1,023,000 What is the decline rate Sales in 5 = 877,225 units And the idea that the market is in decline Sales in 4 = decline rate (Sales in 3). phase of the Product Life Cycle (PLC) 1,023 = decline rate (1,193 ). Do you na ve or last period + decline % Decline rate = 1,023 / 1,193 = 7. Period Actual Units Na ve Forecast error Period Actual Units Na ve Forecast error Period Actual Units Na ve Forecast error Sold Sold Sold A F E=A-F A F E=A-F A F E=A-F. 3 1,193,000 3 1,193,000 3 1,193,000. 4 1,023,000 1,193,000 170,000 4 1,023,000 1,193,000 170,000 4 1,023,000 1,193,000 170,000. 5 ? 5 1,000,000 or the Smallest error 5 1,000,000 or the Smallest error 6 na ve method is naive na ve method is naive 1,023,000 1,023,000.

10 7. 6 6 885,000 977,517 or the Smallest error 7 na ve method is last period +. 1,000,000 decline rate 7. Use last period and decline rate to Use last period and decline rate to Use last period and decline rate to Forecast period 5 Forecast period 5 or na ve method Forecast period 5. 8.


Related search queries