Within this class we offer methods which quantitatively access how well a given probability distribution approximates a given data set.
For a list of all members of this type, see ContinuousProbabilityDistributionApproximation Members.
System.Object
ContinuousProbabilityDistributionApproximation
Further Explanation
In order to access to what degree a given data set can be modeled by a given (continuous) distribution we introduce the squared errror measure of the approaximatiuon error. The spuared error is the sum of the squares of the distances between the frequency of the grouped data set (see details below) of the probability distribution at the midpoints of the grouped data set.
Groupded Data Sets
In order to assist in the analysis of a given data set it is often convenient (and in fact often nessesary) to collect the data into categories. For example, consider a data set which consists of the heights of a set of 100 people. Now in order to derive qualitative properties concerning this data set it is natural to group the data set into height intervals (i.e. such intervals could be 1.20m (or less), 1.21m - 1.50, 1.51- 2.00m, 2.01m (or more)). For each interval the number of members (i.e. frequency) of the original data set which fall within the interval would be recorded.
Remark A number of methods which can be applied in order to evaluated the mean, variance and standard deviation of grouped sets can be found in the Statictics GroupedData class of the Statistics :module:. Note that with this class the midpoint of each data class is treated as though it were the mean of all items in that class.
Namespace: WebCab.COM.Statistics.PDistributions
Assembly: WebCab.COM.StatisticsDemo (in WebCab.COM.StatisticsDemo.dll)
ContinuousProbabilityDistributionApproximation Members | WebCab.COM.Statistics.PDistributions Namespace