Chapter 3 - Demand Forecasting Flashcards
Demand forecast
The estimate of expected demand for a specified period of time in the future.
What are the three types of uses for demand forecasts in operations?
1) to help managers design the system (long-term plans)
2) To help them plan the medium-term use of the system
3) To schedule the short-term use of the system.
What are the long-term plans when designing the system?
- Which goods and services to offer
*Which capacities, facilities and equipment to have
*Where to locate
What are the medium-term plans when designing the system?
*Planning overall inventory and workforce levels
*Planning production at the aggregate product family level
What are the short-term plans when designing the system
*Scheduling of production
*Purchasing of parts and raw materials,
*Staff scheduling
What are the two important aspects of demand forecasts?
1) The expected level of demand (forecast)
2) the degree of accuracy that can be assigned to a forecast
Forecasting horizon
The time period we are forecasting for
what are the elements of a good forecast
1) Timely
2) Accurate
3) Reliable
4) meaningful units
5) in writing
6) simple to understand and use
7) cost-effective
What are the 6 steps to the forecasting process?
1) Determine the purpose of the forecast
2) establish a forecasting horizon
3) gather and analyze relevant historical data
4) select a forecasting technique
5) prepare the forecast
6) monitor the forecast
Judgmental methods
Use nonquantitative analysis of historical data and/or analysis of subjective inputs from consumers, sales staff, managers, executives, similar products, and experts to help develop a forecast
Time series models
identify specific patterns in the data and project or extrapolate those patterns into the future, without trying to identify causes of the patterns
Associative models
use equations that consist of one or more explanatory variables that can be used to predict future demand for the variable of interest.
Delphi method
Involves circulating a series of questionnaires among experts. Responses are anonymous and the new questionnaire is developed using the information extracted from the previous one.
Level (Average)
or constant, refers to a horizontal (flat) pattern of time series. In this setting, the data do not show any particular pattern over time and are, generally, constant.
Trend
refers to a persistent upward or downward movement in the data. Population growth, increasing incomes, and cultural changes often account for each movements.
Seasonality
refers to regular repeating wavelike variations generally related to factors such as the calendar, weather, or recurring events. For example, sales of ice cream are higher in the summer. Restaurants, supermarkets and theatres experience weekly and even daily “seasonal” variations.
Cycles
are wavelike variations lasting more than one year. These are often related to a variety of economic, political, and even agricultural conditions, such as supply of cattle.
Irregular variations
Are due to unusual one-time explainable circumstances not reflective of typical behavior, such as severe weather conditions, strikes or sales promotions. They do not reflect typical behavior, and whenever possible should be identified and removed from the data.
Random variations
are residual variations that remain after all other behaviors have been accounted for (also called noise). This randomness arises from the combined influence of many perhaps a great many relatively unimportant factors and it cannot be reliably predicted
Naive Method
the naive method can be used with a stable series (level with random variations),a simple but widely used approach to forecasting.
Averaging method
Averaging techniques smooth fluctuations in a time series because the individual highs and lows in the data offset each other when they are combined into an average.
What are the three techniques for averaging?
1) Moving Average
2) Weighted moving average
3) Exponential smoothing
Moving average
technique tries to overcome the issue of not having the random variations smoothed out, by expanding the amount of historical data the forecast is based on.
Weighted moving average
similar to a moving average, except that i tallows for the assignment of different weights for different periods included in the forecasting analysis.
Exponential smoothing
a sophisticated weighted averaging method, where the forecast for each period is based on the forecast for the previous period plus a percentage of the forecast error.
trend-adjusted exponential smoothing
variation of simple exponential smoothing can be used when a time series exhibits a linear trend.
what are the two different models of seasonality and what is the differences between the two?
1) additive - seasonality is expressed as a quantity (e.g., 20 units), which is added to or subtracted from the series average in order to incorporate seasonality.
2) Multiplicative model - seasonality is expressed as a proportion of the average (or trend) amount which is then multiplied by the average of the series.
what are the steps to the time series decomposition?
1) compute the seasonal relatives
2) deseasonalize the demand data
3) Fit a model to the deseasonalized demand data
4) forecast using this model (To obtain the deseasonalized forecasts)
5) Reasonalize the deseasonalized forecasts.
least square line
minimizes the sum of the squared deviations around the line.
Correlation Coefficient
measures the strength, as well as the direction, of the relationship between two variables. A positive correlation coefficient indicates that the two variables move in the same direction a negative correlation coefficient means the two variables move in opposite directions.
forecast error
the difference between the value that actually occurs and the value that was forecasted for a given time period.
what are the three alternative forecast-error summaires?
1) Mean absolute deviation (MAD)
2) Mean Squared error (MSE)
3) Mean absolute percent error (MAPE)
Bias
The sum of forecast errors
Positive bias - forecasts frequently underestimate the actual values
Negative bias - forecasts frequently overestimate the actual values
What are some examples of forecast errors?
1) the omission of an important variable
2) a change or shift in the variable that the model cannot deal with
3) The appearance of a new variable.
4) Irregular variations from severe weather or other natural phenomena, temporary shortages or breakdowns, catastrophes, or similar events.
5) random variations randomness is the inherent variation that remains in the data after all causes of variation have been accounted for. There are always random variations.
what are the two most important factors to consider when forecasting?
1) cost
2) Accuracy
