Terms associated with forecasting include:
Descriptive statistics: Forecasting software often must be able to alculate the numerical values representing important features of a set of quantitative information, such as the arithmetic mean, range, standard deviation, ratio, percentage and rate of change.
Regression: Regression is a means for relating one variable to another. Multiple linear regression correlates a variable of interest (the dependent variable) to one or more indepedent variables. Among business applications, this type of program can be used to forecast sales and production levels.
Correlation analysis: A simple method of measuring the strength of a relaionship between two variables.
Moving averages: A forecasting technique, moving averages calculates a moving sequence of data to an arithmetic average.
Exponential smoothing: This is another forecasting technique, based on using weighted sums of past data to forecast a time series.
Census X-11: A sophisticated decomposition forecasting technique developed at the Department of Labor and used for monthly or quarterly data.
Box Jenkins: Combines tentative model identification, estimation and diagnostic checking.
Decomposition method: An approach to forecasting in which a time series consists of a number of unobservable components, such as trend, cycle, seasonality an irregularity.
Automatic forecasting: Refers to the software's ability to automatically analyze historical data and to apply the most effective forecasting methodology to the data.
Override automatic forecasting: When the user knows of certain influencing fctors that the software cannot take into account, overriding the automatic forecasting feature is then necessary to obtain the most accurate forecast.
Bar graph: A graph that illustrates numeric data as a set of evenly spaced bars.
Line graph: A graphic that represents numeric data as a set of points along a line.
Histograms: Histograms are graphical representations of a distribution function by means of rectangles. The width of the rectangle represents intervals into which the range of observed values is divided; the height represents the number of observations occurring in each interval.
Scatter plot: A diagram of corresponding pairs of points that depict a relationship.