3. Forecasting

adaptive (or variable response) exponential smoothing

A version of exponential smoothing where the smoothing constant is automatically modified in order to prevent large forecast errors from occurring.

annual average method

A simpler method for finding seasonal relatives that involves finding the ratio of actual demand relative to average seasonal demand in that year and averaging the ratios across years.

associative models

Use explanatory variables to predict future demand for the variable of interest.

bias

The sum of forecast errors.

centred moving average (CMA)

A moving average positioned at the centre of the data that were used to compute it.

control chart for forecast errors

A time series plot of forecast errors that has limits for individual forecast errors.

correlation coefficient

A measure of the strength of relationship between two variables.

cycles

Wavelike variations lasting more than one year.

Delphi method

Experts complete a series of questionnaires, each developed from the previous one, to achieve a consensus forecast.

demand forecast

The estimate of expected demand during a specified future period.

exponential smoothing

Weighted averaging method based on previous forecast plus a percentage of the difference between that forecast and the previous actual value.

forecast error

Difference between the actual value and the forecast value for a given period.

forecasting horizon

The range of time periods we are forecasting for.

irregular variations

Caused by unusual one-time explainable circumstances not reflective of typical behaviour.

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.

least squares line

Minimizes the sum of the squared deviations around the line.

level (average)

A horizontal pattern of time series.

linear trend equation

ŷ_t = a + bt, used to develop forecasts when linear trend is present.

mean absolute deviation (MAD)

The average of absolute value of forecast errors.

mean absolute percent error(MAPE)

The average absolute percent forecast error.

mean squared error (MSE)

The average of squared forecast errors.

moving average

Technique that averages a number of recent actual values as forecast for the current period. It is updated as new values become available.

naïve forecast

For a stable series, the next forecast equals the previous period’s actual value.

predictor variables

Variables that can be used to predict values of the variable of interest.

random variations

Residual variations after all other behaviours are accounted for (also called noise).

regression

Technique for fitting a line to a set of points.

seasonal relatives

Proportion of average or trend for a season in the multiplicative model.

seasonal variations

Regularly repeating wavelike movements in series values that can be tied to recurring events, weather, or a calendar.

seasonality

Regular wavelike variations related to the calendar, weather, or recurring events.

standard error of the estimate

A measure of the scatter of points around a regression line.

time series

A time-ordered sequence of observations taken at regular intervals oftime

time series models

Extend the pattern of data into the future.

tracking signal

A measure used to control the forecasting process: sum of forecast errors divided by mean absolute forecast error.

trend

A persistent upward or downward movement in data.

trend-adjusted exponential smoothing

Variation of exponential smoothing used when a time series exhibits trend.

weighted moving average

A variation of moving average where more recent values in the time series are given larger weight in calculating a forecast.