Please enable JavaScript.
Coggle requires JavaScript to display documents.
Forecasting demand (Forecasting process (Determine the purpose of forecast…
Forecasting demand
Forecasting process
Determine the purpose of forecast
Determine level of aggregation and what will be forecasted
Determine the time horizon
Visualize the data
Choose the forecasting method or model
Prepare the data
Test the forecast using historical data
Forecast
Perform sales and operations planning
Periodically review and improve models for accuracy
Forecasting methods
Qualitative methods
Judgmental/ expert judgment forecasting
Delphi method
Quantitative methods: time- series forecasting
Visualizing
Deseasonalizing
Calculate the month average for each month
Calculate the year average
Calculate the seasonal index
Simple and weight moving averages, exponential smoothing
simple moving average: 3-month moving average=(M1+M2+M3)/3
The weight moving average: 3-month weighted moving Average=((1
M1)+(2
M2)+(3*M3))/6
Exponential smoothing= (A
last period demand)+[(1-A)
last period forecast]
Reseasonalizing
Quantitative methods: associative forecasting
Correlation versus causation
Leading and lagging indicators
Simple regression
Multiple regression
Service- sector forecasting
Combination methods
Measures of forecast error
Forecast error =Actual-forecast demand
Error as percentage= (A-F)/A
Forecast accuracy =1- forecast error as percentage
Bias and random variation
Cumulative forecast error= cumulative actual demand - cumulative forecast demand
Mean absolute deviation(MAD)=Σ｜A-F｜/n
Tracking signal= Algebraic Sum of Forecast Error/Mean absolute Deviation(MAD)
Forecast bias: if the tracking signal is continually negative, we are consistently over-forecasting
Suitability of the forecasting method:±4 view as working correctly
Standard deviation= MAD*1.25
Mean squared error(MSE)
MSE and MAD comparison
Factors affecting demand
Trends
cycles and other external drivers
seasonality
Promotions or other internal drivers
Random variation
Principles of forecasting
Forecasts are always wrong
Forecasts should include an estimate of error
Forecast are more accurate for groups than single items
Forecasts of near-term demand are more accurate than long-term forecasts