Abstract: Methods and apparatuses for predicting set of multi-dimensional dependent data and non-measurable data from a set of multi-dimensional historical dependent and causal data are described. In one embodiment, the method comprises receiving input data that comprises multi-dimensional historical dependent data and causal data and anticipated activity data, determining a set of multi-dimensional predicted dependent data using a predictive model and the input data, creating non-measurable data based on the set of multi-dimensional predicted dependent data and the input data.

Abstract: Selectivity estimates are produced that meet a desired confidence threshold. To determine the confidence level of a given selectivity estimate for a query expression, the query expression is evaluated on a sample tuples. A probability density function is derived based on the number of tuples in the sample that satisfy the query expression. The cumulative distribution for the probability density function is solved for the given threshold to determine a selectivity estimate at the given confidence value.

Abstract: Methods and apparatuses for updating a forecast model quantifying the marketing of to the demand for a product and/or service are described. An original forecast model created at a reference M is used. An error of the forecast model is determined based on data including an original data and an additional data. At least one parameter is identified to be changed in value in, added to, or removed from the original forecast model. The forecast model is then modified to reduce the error of the forecast model by changing the value of the identified parameter in the forecast model, adding the identified parameter to the original forecast model, and/or removing the identified parameter from the original forecast model.